Journal Issue
Share
Article

Saudi Arabia: Selected Issues

Author(s):
International Monetary Fund. Middle East and Central Asia Dept.
Published Date:
October 2015
Share
  • ShareShare
Show Summary Details

Assessing the Resilience of Saudi Banks to Weaker Economic Conditions1

Banks in Saudi Arabia are profitable, liquid, and well-capitalized. Nevertheless, the recent sharp decline in oil prices is likely to impact the banking system, particularly if it is sustained. Scenario analyses relying on publicly available bank-by-bank data suggest that: (i) banks are well positioned to weather the impact of an increase in nonperforming loans (NPLs), lower profits, and weaker deposit inflows that may come with an extended period of lower oil prices and weaker nonoil GDP growth; and (ii) bank capital and liquidity would only be put under significant pressure in the event of a very sharp economic downturn, sustained very low oil prices, and substantial deposit withdrawals.

A. Background

1. Commercial banks in Saudi Arabia are profitable, liquid, and well-capitalized. On average, NPLs are low at 1.1 percent of gross loans, the capital adequacy ratio is 17.8 percent, and provisions have increased to 183 percent of NPLs.2 Overall, corporate balance sheets are in good shape. The Saudi Arabian Monetary Agency (SAMA)’s regulation and supervision of the banking system has continued to strengthen in recent years, including through the early adoption of Basel III capital and liquidity standards.

2. Historically, NPLs seem to have been influenced by oil prices, government spending, and growth in the non-oil sector (Figure 1). Sustained lower oil prices would over time lead to fiscal tightening and reduce the growth rates of nonoil private sector GDP and real credit extension. As economic activity moderates, equity prices would decline, creating negative wealth effects. As a result, the creditworthiness of borrowers would worsen and liquidity conditions tighten, increasing bank NPLs. Higher interest rates as the U.S. starts to tighten monetary policy would also raise borrowing costs and could put additional pressure on asset quality. On the liquidity side, deposits have historically been correlated with oil prices—lower oil prices reduce income and deposit inflows, or even trigger a draw down, particularly by companies most affected by the price drop.

Figure 1.Key Economic Indicators in Saudi Arabia, 1995–2014

Source: IMF Staff Calculations

3. The 2011 Financial Sector Assessment Program (FSAP) Update looked in detail at the resilience of the banking system to asset quality and liquidity shocks. It concluded that the domestic banking system is generally resilient to a wide range of asset quality and liquidity shocks and would only come under pressure if hit by a severe stress event. Starting from an NPL ratio of 3 percent, provisions covering about 116 percent of NPLs, and a capital ratio of 17.7 percent, the aggregate capital ratio was found to remain above 8 percent for almost all of the individual shocks considered. However, if oil prices remained at $40 a barrel and real GDP growth stagnant for four years, the aggregate capital ratio would decline to 5.2 percent. The liquidity risk assessment in the 2011 FSAP also suggested the resilience of the banking system to most liquidity shocks.

4. This note does not try to update the detailed stress tests conducted in the 2011 FSAP or preempt those that may be carried out in a future FSAP. Rather, it uses publicly available bank-by-bank data, regression analysis, and a range of economic scenarios to revisit the possible impact of lower oil prices and higher interest rates on Saudi banks. The results should be interpreted with a range of caveats in mind. First, the information content of publicly available bank-level balance sheet data is relatively limited compared with the regulatory data typically used for FSAP assessments. Second, any analysis based on historical data might not always account for the effects of recent changes in policy frameworks. Third, the data spanning 1999–2014 may not capture a sufficient number of oil price and financial cycles. Fourth, there is considerable parameter uncertainty surrounding the estimated relationship between macroeconomic shocks and NPL ratios. This is discussed in more detail below.

B. Bank balance sheet quality under weaker economic conditions

5. NPLs in Saudi Arabia appear to be driven by oil prices and nonoil private sector growth. The relationship between macro and financial market variables and NPLs ratios was estimated through panel data econometric techniques. The analysis relied on publicly available bank-by-bank data on balance sheets and profit/loss accounts, focusing on banks for which sufficient data are available for 1999–2014. The econometric approach employed broadly follows Espinoza and Prasad (2010) who analyzed NPLs in the GCC banking system relying on panel data techniques.3 The results suggest that the growth rates of real oil prices and nonoil private sector GDP are key determinants of bank-level NPL ratios (Table 1).4 By contrast, real government spending growth and domestic and U.S. interest rates are not found to directly affect NPL ratios in a systematic way (not shown in the table). When regressed together with the growth rates of real oil prices and nonoil private sector GDP, the coefficients of real equity price growth are statistically significant only in one specification while those on bank-level real credit growth are not statistically significant. The 2008/09 time dummy is significant across all specifications.

Table 1.Determinants of Bank NPLs in Saudi Arabia
Model number123456789
Logit of NPL ratio (L1)0.944***0.911***0.957***0.895***0.845***0.954***0.855***0.852***0.964***
Real oil prices, % change (L1)−0.011***−0.010***−0.011***−0.010***−0.009***−0.011***−0.009***−0.009***−0.011***
Nonoil private sector GDP growth, % (L1)−0.132***−0.127***−0.004−0.113***−0.098**−0.002−0.100**−0.101*−0.001
Real credit growth, % (L1)−0.001−0.001−0.001
Real equity price growth, % (L1)0.0020.0020.003**
2008/09 dummy0.434**0.436**1.045***0.448**0.489**1.048***0.493**0.485**1.061***
Number of observations126126126126126126126126126
Lag depth of GMM instruments111222333
P values
AR(1)0.0120.0120.0240.0120.0190.0290.0230.0250.021
AR(2)0.3170.5910.7740.3440.6090.8300.4610.5110.850
Hansen0.1600.6090.0410.0630.1030.1230.7661.0000.203
Note: Dependent variable is bank-by-bank (logit transformed) NPL ratio for 9 Saudi Arabian banks spanning 1999-2014 (annual frequency). Relying on a system GMM approach. The coefficients represent non-liner effect that depends on starting levels. ***, **, and * signify significance at the 1%, 5% and 10% levels. L1 signifies one period lag. AR(1) and AR(2) signify p-values associated with the null hypothesis of lack of first and second order serial correlation. Hansen signifies p-value associated with the null hypothesis that the instruments are exogenous.Sources: Bankscope, Haver, Bloomberg, and staff estimates.
Note: Dependent variable is bank-by-bank (logit transformed) NPL ratio for 9 Saudi Arabian banks spanning 1999-2014 (annual frequency). Relying on a system GMM approach. The coefficients represent non-liner effect that depends on starting levels. ***, **, and * signify significance at the 1%, 5% and 10% levels. L1 signifies one period lag. AR(1) and AR(2) signify p-values associated with the null hypothesis of lack of first and second order serial correlation. Hansen signifies p-value associated with the null hypothesis that the instruments are exogenous.Sources: Bankscope, Haver, Bloomberg, and staff estimates.

6. Using these parameter estimates and projections of oil prices and nonoil private sector GDP growth, the future estimated path of bank-level NPLs can be derived.5 Based on the estimated relationship, NPL ratios are projected for the individual banks for 2015–19, starting from NPL ratios at end-2014, as new NPLs accumulate according to the baseline trajectories of oil prices and nonoil private sector GDP growth. Oil prices and nonoil private sector GDP growth follow the central projections by IMF staff. In particular oil prices decline from an average of $96 a barrel in 2014 to $59 a barrel in 2015, and recover to $71 a barrel in 2019 (Figure 2, left panel). Nonoil private sector GDP growth moderates from 5.6 percent in 2014, to 3.4 percent in 2015 and 3.8 percent in 2016, respectively, and stabilizes at 5 percent (center panel). Under these assumptions, the aggregate NPL ratio on average rises gradually to 2.8 percent by 2019 (right panel), with the underlying bank level NPL ratios ranging between 2.2–3.2 percent. Using the range of auto-regressive coefficients suggested by the regression results, the aggregate NPL ratio in 2019 could be between 1.3–5.2 percent (right panel).6

Figure 2.NPL Ratio and its Key Determinants: Historical and Scenario I

Source: IMF Staff Estimates.

Note: Oil prices represent the simple average of prices of U.K. Brent, Dubai Fateh, and West Texas Intermediate crude oil. The purpose of this analysis is to explore the resilience of the banking system in Saudi Arabia under alternative macro-financial scenarios. They are not detailed stress tests as conducted in the 2011 FSAP Update, nor do they preempt those that may be carried out in a future FSAP.

7. Given the NPL path, balance sheets and profit/loss accounts are simulated for the individual banks. Liabilities remain constant while interest margins on current loans and liabilities, as well as net non-interest income, decline from each banks’ historical level. This assumption reflects potential margin compression due to slower economic activity, weaker credit demand, and potentially greater competition for funding. New NPLs are assumed to be provisioned at 120 percent, above regulatory requirements of 100 percent. This further dents profits. When the capital ratio declines in the previous period, and provided that net income is positive in the current period, it is assumed that the bank builds capital by allocating 50 percent of profits. The rest is paid out as dividends. When net income is negative, capital covers the loss.

8. Simulation results suggest that banks can comfortably withstand higher NPLs and lower profits under this economic scenario. The finding owes to Saudi banks’ strong starting position, with low NPLs, adequate provisioning, and solid profitability. Based on the abovementioned assumptions and the central path of the NPL ratio, the average capital ratio remains above 18 percent (Table 2, Scenario I). This is despite 120 percent of new NPLs being provisioned, which dents profits but helps maintain provisions at above 140 percent of total NPLs.

9. A second scenario was considered in order to assess the resilience of the banking system to a sharper fall in oil prices and nonoil private sector GDP growth. In scenario II, oil prices are assumed to fall from $96 a barrel in 2014 to $44 a barrel in 2015 and remain little changed, while nonoil private sector GDP growth declines to 0.8 percent in 2015 and 1.3 percent in 2016, before stabilizing at 2.5 percent (these are 1.5 and 1 standard deviations below the levels assumed in scenario I, respectively). Bank profitability moderates further and banks provision 100 percent of new NPLs.

10. The banking system continues to be resilient to the shock in scenario II (Figure 3). NPLs rise to 7 percent and provisions cover about 1.1 times the stock of NPLs. Despite profitability declining and provisioning needs rising, the capital ratio declines only moderately to around 17 percent in aggregate. For one bank, the capital ratio declines below 12 percent, but remains above the 8 percent international regulatory minima. SAMA implicitly sets the regulatory capital minima equal to 12 percent, 4 percentage points above Basel requirements.7

Figure 3.NPL Ratio and its Key Determinants: Historical and Scenarios I-III

Source: IMF Staff Estimates.

Note: Oil prices represent the simple average of prices of U.K. Brent, Dubai Fateh, and West Texas Intermediate crude oil. The purpose of this analysis is to explore the resilience of the banking system in Saudi Arabia under alternative macro-financial scenarios. They are not detailed stress tests as conducted in the 2011 FSAP Update, nor do they preempt those that may be carried out in a future FSAP.

11. A third scenario is constructed to estimate the extent of GDP growth slowdown that would be needed to reduce the average capital ratio to 12 percent in 2019. In scenario III, all assumptions other than the trajectory of nonoil private sector GDP growth remain unchanged. To push the average capital ratio to 12 percent, nonoil private sector GDP would need to contract by 2.6 percent in 2015 and then remain flat on average between 2016–19 (Figure 3). The last time private nonoil growth contracted was in 1987. The NPL ratio would increase to 14.1 percent in 2019 in this scenario according to the regression results. Bank level capital ratios would fall below 12 percent for 8 banks of which 5 banks would maintain capital ratios above 8 percent (Table 2, Scenario III). Resources required to bring the 8 banks’ capital ratios back to 12 percent are small.

Table 2.Effects of Economic Scenarios on the Banking Sector
201420152016201720182019
HistoricalScenario I
Assum ptions
Nonoil private sector growth (percent)5.63.43.84.75.05.0
Oil prices (US dollars)965964677071
Impact
Nonperforming loans (percent of total loans)1.11.32.52.72.82.8
Provisions (percent of NPLs)182.9173.2148.4146.2145.3145.2
Capital adequacy ratio17.818.518.618.618.618.6
CAR < 8 percent
Number of banks000000
8 percent < CAR < 12 percent
Number of banks000000
CAR > 12 percent
Number of banks121212121212
HistoricalScenario II
Assumptions
Nonoil private sector growth (percent)5.60.81.32.12.52.5
Oil prices (US dollars)964443444444
Impact
Nonperforming loans (percent of total loans)1.11.33.75.26.27.0
Provisions (percent of NPLs)182.9170.1125.3117.8114.9113.2
Capital adequacy ratio17.818.418.017.817.416.9
CAR < 8 percent
Number of banks0.000000
8 percent < CAR < 12 percent
Number of banks0.000011
CAR > 12 percent
Number of banks12.01212121111
HistoricalScenario III
Assumptions
Nonoil private sector growth percent5.6−2.6−1.30.00.50.5
Oil prices (US dollars)964443444444
Impact
Nonperforming loans (percent of total loans)1.11.35.18.711.614.1
Provisions (percent of NPLs)182.9170.1118.3110.7108.0106.5
Capital adequacy ratio17.818.417.115.413.812.0
CAR < 8 percent
Number of banks0.000013
8 percent < CAR < 12 percent
Number of banks0.000245
CAR > 12 percent
Number of banks12.012121074
Note: Oil prices represent the simple average of prices of U.K. Brent, Dubai Fateh, and West Texas Intermediate crude oil. Nominal oil prices are deflated by inflation for regression and scenario analysis. The purpose of this analysis is to explore the resilience of the banking system in Saudi Arabia under alternative macro-financial scenarios. They are not detailed stress tests as conducted in the 2011 FSAP Update, nor do they preempt those that may be carried out in a future FSAP.
Note: Oil prices represent the simple average of prices of U.K. Brent, Dubai Fateh, and West Texas Intermediate crude oil. Nominal oil prices are deflated by inflation for regression and scenario analysis. The purpose of this analysis is to explore the resilience of the banking system in Saudi Arabia under alternative macro-financial scenarios. They are not detailed stress tests as conducted in the 2011 FSAP Update, nor do they preempt those that may be carried out in a future FSAP.

12. The impact of the decline in oil prices on bank deposits is likely to be manageable. Historically, bank deposits have been correlated with oil prices (and with the NPL ratio). Results of a panel vector auto regression using annual bank-by-bank data for 1999–2014 suggest that the growth rate of deposits in real terms slows by 0.1–0.2 percent in response to a one percent decrease in real oil prices, or by 1-2 percent in response to a one percentage point increase in the NPL ratio (see IMF (2015)). Taking the results at face value, the recent 40 percent decline in oil prices, if sustained for one year, would lead to a 4-8 percentage point reduction in deposit growth, everything else constant. The ability of Saudi banks to manage a much larger deposit withdrawal is discussed in Box 1.

Box 1.The Resilience of Saudi Banks to Deposit Withdrawals

Liquidity positions in the Saudi banking system are well managed. Banks hold sufficient high-quality liquid assets to already bring both the Liquidity Coverage Ratio (LCR) and the Net Stable Funding Ratio (NSFR) above the regulatory minimum requirements proposed by Basel III (see SAMA (2015)).

The impact of more severe scenarios where banks actually experience significant deposit withdrawals can be considered. The extent to which banks can accommodate deposit reductions by selling liquid (and to some extent illiquid) assets within a 30-day window (divided into five periods, six days each) can be investigated. In the scenarios, withdrawal rates vary by the type of deposits, with demand deposits suffering higher runoff rates. During times of market stress, banks would not be able to convert all their liquid assets into cash at face value. Moreover, some liquid assets are encumbered in margin calls–a higher amount of assets are encumbered during a more severe scenario to meet greater collateral demand. Two scenarios are parameterized to calibrate Saudi Arabia’s historical experience, guided to some extent by Schmieder et al. (2012) (Figure).

  • In scenario A, demand and time deposits decline by 2 percent and 1 percent per period, respectively, or by an average of 8 percent over the 30 day period. This is higher than the largest monthly reduction observed since 1993 which was 6 percent. 95 percent of liquid assets are available for sale in each 5-day period (2 percent for illiquid assets). Liquid assets are sold with a moderate 1 percent haircut (15 percent for illiquid assets) and 10 percent of liquid assets are encumbered (25 percent of illiquid assets), both reducing the capacity of banks to generate cash.

  • Scenario B is characterized by a faster deposit run and tighter market liquidity conditions. Deposits decline by more than 10 percent during a 30 day window, a rate similar to the 11 percent reduction within a week suffered in 1990 (when the run was triggered by a military conflict rather than an oil price decline). Smaller shares of assets are available for sale (85 percent for liquid assets and 1 percent for illiquid assets). Banks face higher rates of haircut (5 percent and 30 percent) and encumbrance (20 percent and 50 percent).

Net Cash Flow

(Percent of initial total assets)

Source: IMF Staff Estimates.

Note: The purpose of this analysis is to explore the resilience of the banking system in Saudi Arabia under alternative macro-financial scenarios. They are not detailed stress tests as conducted in the 2011 FSAP Update, nor do they preempt those that may be carried out in a future FSAP.

The results suggest that the banking sector is generally resilient to the deposit withdrawals, although some liquidity shortfall emerges in scenario B.1 In scenario A, there is no liquidity shortfall for the banking sector as a whole or for any individual bank. In scenario B, the aggregate banking system remains liquid, but five banks suffer liquidity shortages by 1–2 percent of initial assets (banks #3,4,7,8,10). Similar to the asset quality scenarios, the assessment needs to be interpreted with the range of assumptions in mind.

1 This analysis assumes lower rates of deposit runoff, haircut and encumbrance than those suggested in Schmieder et al. (2012).
References

    EspinozaRaphael and A.Prasad2010Nonperforming Loans in the GCC Banking System and their Macroeconomic Effects,IMF Working Paper 10/224International Monetary FundWashington.

    International Monetary Fund (IMF)2014Assessing Concentration Risks in GCC Banks,Prepared for the Annual Meeting of Ministers of Finance and Central Bank Governors in Gulf Cooperation CouncilInternational Monetary FundWashington.

    International Monetary Fund (IMF)2015Assessing the Importance of Oil and Interest Rate Spillovers for Saudi Arabia.Saudi Arabia Article IV Consultation Selected IssuesIMF country ReportWashington.

    Saudi Arabian Monetary Agency (SAMA)2015Financial Stability ReportRiyadh.

    SchmiederC.H.HesseB.NeudorferC.Puhr and S.Schmitz2012Next Generation System-Wide Liquidity Stress Testing.IMF Working Paper 12/3International Monetary FundWashington.

Prepared by Ken Miyajima.

Based on the publicly available bank-by-bank data used in this note.

The NPL ratio exhibits a strong autocorrelation and the data’s time series dimension is short relative to its cross-sectional dimension. This argues for a GMM estimation approach rather than a fixed effects approach–the latter suffers from a downward Nickell bias in such circumstances. Indeed, the coefficient on the autoregressive NPL ratio is smaller when estimated using a fixed effects approach. This would make the trajectory of projected NPL ratios higher compared to the results reported in this note.

NPL ratios are introduced after a logit transformation. Fisher-type panel unit root tests reject the null hypothesis that all panels contain unit roots. Time dummy variables were introduced in the regressions to control for events other than oil price developments that potentially led to an increase in NPL ratios around the time of the global financial crisis. In particular, two large family-owned conglomerates defaulted on loans in 2009 due to events unrelated to the decline in oil prices.

Real oil prices used for regression and projection are converted to nominal prices for discussion. Using the simple average of prices of U.K. Brent. Dubai Fatch, and West Texas Intermediate crude oil.

The range is estimated considering only the coefficients on the autoregressive term which has a large impact on the projected path of NPL ratios. Other sources of uncertainty include the estimated coefficients Following Espinoza and Prasad (2010), bank by bank variables are considered as endogenous and aggregate variables predetermined in this analysis.

Staff estimates suggest that for banks in Saudi Arabia the minimum capital requirement after accounting for concentration risk could reach 12–13 percent of risk-weighted assets. See IMF (2014).

Other Resources Citing This Publication