6. How Might the Sustained Decline in Oil Prices Affect MENA and CCA Banking Systems?
- International Monetary Fund. Middle East and Central Asia Dept.
- Published Date:
- October 2015
The slump in oil prices, through its adverse impact on oil-dependent economies, has raised questions about financial sector stability in MENA and the CCA. The risks are more pronounced in the CCA and non-GCC oil exporters, where the impact of the oil price shock has been compounded by spillovers from Russia and other shocks, against the backdrop of already elevated bank vulnerabilities. As low oil prices persist, some banks may become distressed, especially in countries where space for countercyclical policies is limited and/or regulatory and supervisory frameworks are weak. Maintaining sound macroeconomic policies, increasing supervisory oversight, strengthening prudential and crisis management frameworks, and reducing bank vulnerabilities, particularly dollarization, are key to mitigating financial stability risks.
The decline in oil prices has important implications for the MENA and CCA economies and their financial sectors. In oil-exporting countries, lower oil prices are weakening the balance sheets of oil companies and governments and raising credit and liquidity risks for banks through their adverse impact on the broader economy. In countries such as those of the GCC, where the government or oil companies have majority ownership stakes in rated banks, lower oil prices could also undermine the intrinsic strength of banks and raise funding costs for those that tap international markets. In oil importers, lower oil prices have a positive impact on the economy but adverse spillovers from oil-dependent trading partners can partly offset the benefits. Other concurrent shocks, such as intensifying conflicts in non-GCC MENA or spillovers from Russia to the CCA, add to the impact of lower oil prices.
This chapter discusses the financial stability impact of a sustained decline in oil prices for MENA and CCA countries, as well as policies to mitigate macrofinancial risks. It identifies key transmission channels, vulnerabilities, and feedback loops that can amplify the impact of the oil price decline on the banking systems, as well as data gaps that can impede effective financial sector surveillance. It also discusses policy options to mitigate the macro-financial risks for banks. The analysis covers 21 countries, including 14 oil exporters and seven oil importers. Libya and Sudan are excluded because of data limitations. Only oil importers whose major trading partners are net oil exporters are covered.
How Can Low Oil Prices Affect Banking Systems in MENA and the CCA?
In the GCC, slowing government spending presents a major risk for banking systems but it has so far been contained. Government infrastructure spending drives non-oil GDP growth and bank lending to public sector entities and private contractors, whose performance, in turn, affects banks’ credit risks. Bank lending to households is driven by growth in the public sector wage bill. Because most GCC countries have large buffers, slowdowns in government spending, in response to lower oil prices, are expected to be gradual, limiting credit risks (see Chapter 4). Moreover, prudential frameworks have been strengthened to comply with Basel III rules. A lingering concern, though, is that credit risk can be amplified by high loan concentrations to single borrowers and/or sectors, particularly those that are cyclically sensitive, like real estate and construction (Figure 6.1).1 Exchange rate pegs are perceived as credible, thus exchange rate risks are muted.
Figure 6.1MENA and CCA: Banking System Vulnerabilities
Sources: National authorities; and IMF staff estimates.
Note: OE = oil exporters; OI = oil importers. Country abbreviations are International Organization for Standardization (ISO) country codes.
1Includes only lending exposure. Indirect exposure via collateral or investments is not taken into account.
2Includes government and other public sector deposits.
Risks to financial stability are higher in non-GCC MENA oil exporters. Significant bank vulnerabilities remain, while capacity to mitigate the risks is limited because of generally smaller, or inaccessible, buffers and weaker, or absent, macroprudential and crisis management frameworks. In Algeria and Iraq, bank dependence on oil-related deposits (Figure 6.1) and exposure to state-owned enterprises (SOEs), whose performance is driven by oil, against the backdrop of weak corporate governance of both banks and SOEs, increase both credit and liquidity risks. State influence in Iran’s banking system tends to weaken underwriting standards, which puts asset quality at risk. Banking sectors in Iraq and Yemen are also exposed to sovereign credit and liquidity risks from excessive credit exposures to oil-dependent governments, whose fiscal positions have weakened. Gaps in prudential frameworks limit the scope for mitigating these risks.
In MENA oil importers, banking systems are exposed to oil price shocks through their links to oil exporters, particularly the GCC. Remittances from the GCC (see Box 2.1) support liquidity in the banking systems and foreign exchange markets, especially in Jordan and Lebanon, and, to a lesser extent, Egypt; the latter also receives sizable official grants from the GCC. Significant dollarization (Figure 6.1), elevated nonperforming loans (NPLs), and high bank exposure to sovereign debt could increase financial stability risks in the event of a sharp slowdown in the GCC economies. Exposure to the cyclically sensitive real estate and construction sectors is also significant for some countries. Gaps in prudential frameworks heighten the risks.
The CCA banking systems are affected through multiple channels. The impact from lower oil prices, compounded by spillovers from Russia’s slowdown, exchange rate depreciation (see Chapters 3 and 7), and increases in interest rates in response to rising inflation in some countries, has not only increased credit and liquidity risks but also exchange rate and solvency risks for the CCA banks. Banks’ funding strategies—based on intermediation of dollar deposits and foreign currency lending to unhedged borrowers—heighten these risks. In addition, rapid private sector credit growth in the years prior to the recent oil price shock—in Armenia, Azerbaijan, Georgia, the Kyrgyz Republic, and Tajikistan—has increased the likelihood of asset quality deterioration in slowing economies.2
Weak corporate governance in banks and recipient SOEs increase credit risks (Tajikistan). Delays in resolving past NPLs and problem banks also aggravate stability risks in a number of countries (Azerbaijan, Kazakhstan, Tajikistan). Moreover, important gaps in prudential and crisis management frameworks could limit scope for the orderly resolution of problem banks, if the risks were to materialize.
How Has Bank Soundness Been Affected So Far?
Recent financial soundness indicators (FSIs) point to a weakening in bank soundness of several CCA countries (see Figure 6.2).3 NPLs are trending up, profitability has declined, and, although capital adequacy ratios (CARs) remain high, they are declining in most countries.4 Open foreign exchange positions have widened, thus exchange rate depreciations have consequently increased revaluation losses and capital erosion, in addition to indirect credit risks from borrowers in foreign currency. Private sector credit growth has also weakened across the CCA, particularly in real dollar terms.
Figure 6.2CCA: Recent Developments in Banking System Soundness and Performance
Sources: National authorities; and IMF staff estimates.
Note: Country abbreviations are International Organization for Standardization (ISO) country codes.
Contemporaneous aggregate indicators may understate the extent of deterioration in bank soundness. Recent economic shocks are likely to be reflected in the NPL numbers with a lag. Some banks have been restructuring loans (Armenia, Azerbaijan, Kazakhstan). The strength of bank balance sheets is overstated by inadequacies in loan classifications and provisioning (Azerbaijan, Kazakhstan, Tajikistan) and one-off charges to NPL stocks.5 Bank-by-bank analysis also shows a dispersion in bank performance.6
Exchange rate depreciations have had a particularly quick and profound impact on banking system soundness, owing to significant dollarization in the balance sheets of banks and borrowers. Although devaluations helped preserve international reserves and improve fiscal positions, actual and expected devaluations precipitated deposit dollarization, while reducing demand for foreign currency loans (Figure 6.2). Widening currency mismatches between banks’ assets and liabilities are increasing revaluation losses, which erode banks’ capital and constrain local currency loans in the absence of available hedging instruments. Exchange rate depreciation has also increased indirect credit risk among borrowers in foreign currency. Rising deposit dollarization, deposit flight—and policies to either stem currency depreciations or inflation—tightened local currency liquidity in a number of countries (Armenia, Kazakhstan).
Policy responses have aimed to balance the goals of facilitating economies’ adjustment to large external shocks and preserving financial stability (see Box 3.2). In addition to intervention or administrative measures to moderate exchange rate pressures, several countries (Armenia, Azerbaijan, Kazakhstan) have provided liquidity support to banks amid deposit volatility, tried to ease overall liquidity conditions through reduced reserve requirements (Azerbaijan), or placed government deposits and foreign exchange deposits at commercial banks (Tajikistan). Other measures have included the use of foreign exchange swaps to hedge tenge deposits (Kazakhstan); increasing foreign exchange reserve requirements to address rising deposit dollarization and increased capital requirements for banks (Armenia); and reducing loan-loss reserve requirements for restructured loans (Azerbaijan).7
Banking systems in MENA have been more resilient, in aggregate, but there is considerable heterogeneity across countries in bank performance and vulnerabilities (Figure 6.3). CARs remain high and NPLs are low, with the exception of countries whose elevated NPL levels predate recent shocks. Deposit growth in oil-exporting countries (Algeria, Iran, Oman, Qatar,
Figure 6.3MENA: Recent Developments in Banking System Soundness and Performance
Sources: National authorities; and IMF staff estimates.
Note: OE = oil exporters. Country abbreviations are International Organization for Standardization (ISO) country codes.
Saudi Arabia, United Arab Emirates) has begun to moderate, yet bank liquidity remains high. Credit growth is slowing, however, except in Qatar where investment in the run-up to the 2022 FIFA World Cup is driving credit demand.
GCC banking sectors have continued to perform strongly, reflecting solid economic fundamentals and low bank vulnerabilities. Although the oil price shock has eroded fiscal and external surpluses, the impact on economic activity has been limited because large financial buffers have allowed governments to avoid sharp cuts in public spending, supporting consumer and investor confidence and moderating equity price declines (see Chapters 1 and 4). Lending to households is predominantly to public sector employees, whose incomes have not been affected by the decline in oil prices. Banks have benefited from abundant retail deposits while available financing has contained governments’ drawdowns of bank deposits.
In non-GCC MENA oil exporters, the banking sector performance has been mixed, reflecting structural vulnerabilities that predate the oil price shock. Algeria’s exchange rate has depreciated and the economy has slowed, but controls on banks’ foreign exchange exposures and administrative restrictions on lending to households have muted exchange rate and credit risks for banks. Strains in Iran’s banking system have emanated from the effects of sanctions and bank governance issues, while the impact of low oil prices has been less apparent. In Iraq, the economic slowdown and the fiscal crisis, stemming from low oil prices and the insurgency by the Islamic State of Iraq and the Levant (ISIL), increased financial stability risks as banks’ financing of fiscal operations rose. In Yemen, low oil prices, together with intensified conflicts, have weakened the fiscal position and heightened sovereign credit and liquidity risks for banks, because of sizable exposures to government paper.
Banking systems in MENA oil importers have benefited from recent improvements in economic performance. Lower oil prices have reduced fiscal pressures while continued growth of public spending in the GCC has helped sustain remittance inflows and support bank liquidity.
How Vulnerable Are MENA and CCA Banks to Sustained Low Oil Prices?
With low oil prices expected to persist, the economic environment facing CCA and MENA banks will remain challenging. Banks derive most of their income from the domestic market and from lending to households and the non-oil sector (Figure 6.4). A sharper economic slowdown is thus likely to increase credit risks. These risks can be amplified by sectoral (real estate, construction) and single-borrower loan concentrations. A further decline in oil prices could also slow growth in deposits and private sector loans, even if liquidity risks are moderated through central bank facilities.
Figure 6.4Banks Derive Most of Their Income Domestically
Sources: Audited bank financial statements.
Note: Includes data on the largest 60 banks from 11 countries. Domestic income for some GCC countries (Bahrain, Kuwait) likely reflects income from activities in other GCC economies.
Country-specific econometric analyses confirm the strong relationship between oil prices and bank performance in MENA and the CCA. Though the econometric analysis is constrained by the availability of data, GDP growth is consistently found to have the largest impact on NPL growth. Oil prices affect NPLs mostly through GDP and, in some cases, other economic variables, such as exchange rates (Duma 2015). The impact occurs with significant lags but is persistent (Espinoza and Prasad 2010).
In the CCA, exchange rates are an important determinant of NPLs. Interest rates are also important for Georgia, while inflation tends to be a significant determinant of NPLs in Azerbaijan. In Tajikistan, where the economy is highly dependent on remittance inflows from Russia, the corporate sector poses greater credit risk than households, though declines in remittances do have a significant impact on bank asset quality. Remittances from Russia help explain the dynamics of real GDP growth in remittance-dependent countries such as Tajikistan.8
For some MENA banking systems, external financial linkages are an important channel for transmission of shocks. Among GCC countries, because Bahrain’s banks include not only retail banks but also wholesale banks, the broader geographical footprint of the latter group of banks reduces the impact of domestic GDP growth on NPLs (Blotevogel and Sidahmed 2013). In MENA oil importers (Egypt), capital inflows tend to affect asset quality of banks, confirming the importance of external financial linkages (Love and Ariss 2013).
These findings suggest that, although increased stability risks are, at present, pronounced mainly in the CCA, over the longer horizon MENA banking systems may not be immune either. The lags and persistence with which economic slowdowns affect credit risks suggest that macro-financial spillovers from low oil prices may not have played out fully yet and a further deterioration in credit quality is possible.
For CCA banks, the susceptibility to sustained low oil prices has been further increased by the weakening in the balance sheets of banks and borrowers, the nonlinear effects of the macroeconomic shocks on banks, lower buffers to lean against the wind—particularly to mitigate liquidity risks in dollarized banking systems—and gaps in supervisory frameworks. Banks will also face an increasingly challenging operating environment, owing to the effect of slowdowns in domestic economies and in key trading partners such as Russia and China.
GCC banking systems are starting from a position of strength, including in macro-prudential policies and oversight of banks, but a sustained period of low oil prices could increase risks to financial stability if public investment is scaled back sharply or if real estate prices decline. Negative feedback from the banking system to the real economy, through declining credit growth, is also possible as liquidity conditions tighten, because oil-related deposits are a key source of bank funding (see Figure 6.1).
In non-GCC MENA oil exporters, the dominance of state-owned banks, which are dependent on oil-related deposits and exposed to SOEs, increases systemic banking risks (Algeria, Iraq). Rising fiscal pressures aggravate these risks (Iraq, Yemen).
In MENA oil importers (Egypt, Jordan, Lebanon), slower GCC growth could affect remittances and bank deposits, with spillovers to bank credit and foreign exchange markets. Banking stability risks would rise if exchange rates were to come under pressure, given moderate dollarization.
Stress tests highlight similar differences in the resilience of banking systems between MENA and the CCA, as well as the dispersion in risks at the bank level. Generally, credit risk constitutes the single most important risk for banking systems, particularly in the CCA, where it is amplified by exchange rate, interest rate, and concentration risks. IMF Financial Sector Assessment Programs (FSAPs) for Azerbaijan, Georgia, Kazakhstan,9 and Tajikistan, and stress scenarios by country authorities (Kyrgyz Republic) indicate that although in aggregate the banking systems exhibit resilience, adverse shocks can leave a number of banks undercapitalized. Funding risks related to dollarization and, in some cases, reliance on nonresident deposits also present risks.
Stress tests performed during Article IV consultations, and by country authorities in the GCC (Kuwait, Oman, Qatar, Saudi Arabia, United Arab Emirates), indicate that strong capital buffers and low NPLs provide a substantial cushion, though selected banks are vulnerable to severe downside shocks. Overall, recapitalization needs were higher for the CCA than MENA countries.10
Besides the direct impact on financial stability, a sustained decline in oil prices could trigger negative feedback loops between the banking sector and the economy, both in the CCA and MENA. Rising fiscal deficits financed through zero risk—weighted domestic government bonds provide investment opportunities for banks and can have a positive effect on their capital. However, in countries with low excess reserves, government bond financing can crowd out the private sector and accelerate a slowdown in credit as banks become increasingly averse to credit risk in a slowing economy.
How Can Policies Help Mitigate Financial Stability Risks?
Sound macroeconomic policies and increased supervisory vigilance are key to reducing financial stability risks in the CCA and MENA. Low oil prices affect financial stability mainly through their impact on the broader economy; thus, macroeconomic policies that engender growth also help promote financial stability. These should be complemented by enhanced surveillance of credit, liquidity, and solvency risks and regular stress testing. Data gaps should be closed to ensure effectiveness of surveillance, prudential measures need to be strengthened, and preparedness for dealing with bank distress improved. Forbearance should be avoided and shareholders should be called upon to provide capital where needed.
Given the significant differences in financial vulnerabilities across MENA and the CCA countries, policy priorities need to be tailored to country-specific circumstances.
In the CCA, reducing dollarization and strengthening prudential and crisis management frameworks are critical. These need to be supported by steps to address directed lending and improve corporate governance. Reducing dollarization requires tackling its root causes—improving policy credibility and developing financial markets—in addition to differentiating capital requirements for lending to unhedged borrowers (Ben Naceur, Hosny, and Hadjian 2015). To avoid forbearance, restructured loans should be adequately provisioned and open foreign exchange position limits enforced. Gaps in data for macro-financial risk analysis should be addressed.
In the GCC, the liquidity implications of low oil prices and the differential impact of slowing growth on Islamic and conventional banks warrant attention. Coordination between the central bank and the government in financing government deficits can help minimize potential liquidity shocks. Issuance of domestic government bonds would provide compensatory investment opportunities for banks in a slowing economy, and balancing the composition of issuance between conventional bonds and Sukuk could help level the playing field for conventional and Islamic banks. Large exposures of banks to real estate suggest the need to develop metrics that can more comprehensively capture risks to real estate and facilitate the implementation of macroprudential policies. Macroprudential tools should also be expanded to enhance the resilience of banks, in particular, to cyclical risks (Arvai, Prasad, and Katayama 2014).
In non-GCC MENA oil exporters, the priorities include strengthening prudential and corporate governance frameworks and reducing private sector crowding out. There is an urgent need to introduce macroprudential policies and crisis management frameworks, strengthen microprudential regulation and supervision, improve corporate governance, for both banks and SOEs, and close broad-based data gaps.
In MENA oil importers, a combination of macroeconomic policies and supervisory measures is key to minimize stability risks. In particular, there is a need to address vulnerabilities related to dollarization, exposures to government debt, weak asset quality, and inadequacies in prudential frameworks.
Prepared by Inutu Lukonga (team lead) and Moez Souissi with input from Kusay Alkunaizi, Kay Chung, Pritha Mitra, Rafik Selim, Saad Quayyum, Andre Santos, and Bruno Versailles. Research support was provided by Mark Fischer, Gregory Hadjian, Brian Hiland, and Jonah Rosenthal.
Indirect exposures to real estate through collateral and the growing Islamic banking sector’s investments remain high. Available metrics do not capture indirect exposures to real estate, especially in countries with a significant presence of Islamic banks, because banks are permitted to establish subsidiary companies for investment purposes.
Growth in credit, to some degree, reflects currency valuation effects, which implies that the debt burden of foreign currency borrowers has increased substantially, raising the probability of default.
In some cases differences in definition and/or measurement of FSIs affect their comparability across countries.
Banking systems in Uzbekistan and Turkmenistan, where economic growth has remained strong, appear more stable.
In Kazakhstan, bank loans are mostly collateralized by real estate, but banks do not revalue collateral in a timely manner, and estimates are generally based on past high prices. Also, the recent decline in NPLs reflects the revocation of BTA Bank JSC’s banking license and the removal of tax, accounting, and other legal obstacles to write-offs and transfers to special-purpose vehicles, rather than improving asset quality. In Azerbaijan, NPLs are underestimated because only the overdue portions of principal and interest, and not the full amount of the loan, are included in NPL numbers.
In Azerbaijan and Kazakhstan, capital in some banks has fallen below the statutory minimum, while in Armenia, the erosion of bank capital has been moderated by injections of new capital raised to comply with the new minimum statutory capital requirements. In Tajikistan, the placement of government deposits and National Bank of Tajikistan foreign exchange deposits at commercial banks throughout 2014 supported the liquidity of several banks (see IMF Country Reports 15/241 and 15/65; and Press Releases 15/265 and 15/268).
Reducing provisions for restructured loans could have the unintended consequence of encouraging banks to renegotiate loans with borrowers instead of recognizing new NPLs.
For Kazakhstan, assessments of macro-financial risks were updated as part of the recent Article IV consultation.
Stress tests for non-GCC MENA countries have not yet been completed, but the analysis of available bank data suggests that banks could be vulnerable to credit and liquidity risks and recapitalization needs could be substantial. Contingent fiscal liabilities could also increase in countries where state-owned banks are prevalent. Countries with large government exposures also face sovereign risks, and their capital buffers are overstated by the zero risk weights for government securities.