III. Bank Risks from Cross-Border Lending and Borrowing in Slovenia15
Slovene banks are increasingly relying on foreign funding to finance credit, mostly on loans from EU banks. At the same time, to boost profits, banks are increasing their exposure into regions with wider margins, such as the rest of former Yugoslavia.
The expansion of cross-border lending and borrowing can make Slovene banks more vulnerable to interest rate and funding risks, while lending in riskier countries can raise credit and currency risks.
Stress tests do not point to high vulnerabilities, given the still small share of foreign assets, but information on foreign credit exposure could be improved. Future developments should be closely monitored, and stress tests should include more disruptive scenarios of combined shocks.
34. Slovene banks have increased cross-border borrowing and lending to compensate for pressures on profitability at home. Profitability has been under pressure since EU integration and euro adoption, as greater competition is narrowing interest margins and revenues from foreign exchange transactions are being reduced. To increase revenues, Slovene banks have started to expand cross-border lending and capital investments in high-margin regions, such as Southeastern Europe. At the same time, to finance growing demand for credit, while depositors are shifting to higher yielding mutual funds, Slovene banks have resorted to foreign borrowing.
35. These trends can increase vulnerability to various risks. The foreign operations expose Slovene banks to risks related to changes interest rates, direct or indirect currency and credit risks from exposure in riskier markets. The dependency on foreign loans could also increase funding risks, if foreign banks suddenly reduce lending in response to a common shock to the region.
36. The paper assesses the nature of these vulnerabilities using stress tests. After a brief description of the extend of Slovene banks’ cross-border transactions, the paper, based on stress tests conducted by the Bank of Slovenia, analyzes banks’ exposure to interest rate, credit, currency, and liquidity risk. It concludes with some policy observations.
37. The paper finds that vulnerabilities are contained mainly due to the still modest foreign exposure in total assets. The analysis shows that moderate interest rate and margin shocks are not found to put unsustainable pressure on the banking system, as foreign borrowing has reduced the maturity mismatch and lowered the interest rate sensitivity. while the introduction of the euro has lowered currency risks, increasing exposure to foreign credit risk raises vulnerabilities with regard to a downturn in Southeastern Europe (SEE). while these risks warrant greater vigilance, especially as they are concentrated in a few banks, the current size of foreign lending relative to total assets is still small. Liquidity risks, for example, from the withdrawal of foreign funding, are limited for domestic-owned banks, while foreign-owned banks are heavily dependent on financing from their mother institution.
B. Extent of Cross-Border Lending and Borrowing
38. The increase in cross-border assets has been concentrated in a few Slovene banks. While part of this reflects a rise in holdings of European securities after Slovenia’s entry into the EU in 2004, loans to non-residents increased from 6 percent in 2004 to about 8 percent of banks’ balance sheets in 2006. The strongest growth took place in the rest of former Yugoslavia, which reached 3.3 percent of banks’ balance sheets in 2006, and was driven by the largest bank in Slovenia. These exposures include loans, off-balance sheet liabilities and securities. While the exposure to nonresidents is dominated by lending to foreign banks, claims to the foreign nonbanking sector are also increasing. The largest bank has also raised its foreign assets via acquisitions, which will add to the risks from cross-border activities (Box 1). However, this study only covers direct loan growth abroad in the assessment of vulnerabilities.
|Private sector credit||2.3||0.5||2.4||4.7||7.6|
|Other financial institutions||0.0||0.3||0.3||0.2||1.1|
|Money market instruments||0.5||0.6||0.3||−0.8||−0.3|
|Other financial institutions||0.2||0.2||0.4||0.2||0.0|
|Annual GDP (Billions of SIT)||4,800||5,355||5,814||6,272||6,620|
|Total bank assets (percent of GDP)||83||80||82||85||100|Box 1.Foreign Expansion of Nova Ljubljanska Banka (NLB)
The foreign expansion of the banking system is led by NLB, the largest Slovene banking group. NLB accounts for 40 percent of total Slovene banking sector assets, and is majority owned directly and indirectly by the state. Two stylized facts on its foreign expansion stand out--foreign asset exposure through subsidiaries in South-Eastern Europe and Switzerland more than doubled over 2000-05, and it earns a disproportionate share of income and profits from its foreign expansion. for example, assets in Macedonia, Serbia, Bosnia and Herzegovina, and Switzerland grew from 3 percent in 2000 to 8 percent of total in 2005. Over the same period, as a proportion of group profits, contribution of the four countries grew from 1 percent to 19 percent. A similar story emerges for operating income and net interest revenue.
Total Operating Income
Source: Bankscope, staff calculations.
Net Interest Revenue
39. Most banks have increased borrowing from foreign sources. Foreign loan exposure of Slovenian banks grew from 2 to 12 percent of GDP between 2002-05. The annual growth of over 50 percent in recent years has been higher than the average for other new EU members. This helped finance domestic credit growth at around 20 percent annually. Liabilities to foreign banks now account for 30 percent of total liabilities, and are highest in the foreign owned banks.
Net Interest Margin
Sources: Bankscope; and staff estimates.
40. The main forces driving Slovene banks’ foreign activities are wider margins and lower cost of funds. Margins in Slovenia are gradually converging to euro area levels. In 2006, the net interest margins in domestic banks were 2.3 percentage points compared to 1.8 in foreign-owned ones. The other SEE countries, however, offer significantly higher, albeit volatile margins that can go up to 7 percentage points. Expansion to these higher margins markets has helped Slovene banks sustain profits, that have been under pressure in recent years. Reliance on foreign credit, in turn has been less costly that increasing domestic deposit rates, that compete with returns to mutual funds.
|Regulatory capital to risk-weighted assets||11.5||11.8||10.6|
|Net interest margin to average interest bearing assets||3.2||2.8||2.5|
|Return on average assets (before tax)||1.0||1.1||1.0|
|Average short-term assets to average short-term liabilities||93.2||88.4||84.8|
|Variable rate contracts (share of new loans of largest 8 banks)||30.1||38.5||54.5|
C. Stress Tests with Risks from Cross-Border Finance
41. The vulnerabilities related to foreign expansion are assessed using stress tests on shocks to interest rate, credit, currency and liquidity risks. The tests are based on model results provided by Bank of Slovenia (BoS). In some cases, given the differences in foreign activities of Slovene banks, they are divided into three groups in these tests: large domestic (six), small domestic (seven) and foreign owned banks (nine with foreign controlling stakes). Details on the methodology and assumptions are in the Annex.
Interest rate risk
42. Foreign lending and borrowing affect interest rate risk through interest margins and maturities. Interest income can be affected by divergences in foreign lending and borrowing rates. By relying on cheap funding from the euro area and investing in high margin areas, such as SEE, Slovene banks become subject to increased earnings risk from shifts in interest rates. Foreign lending and borrowing can also change the maturity composition on the banks’ balance sheets with different sensitivities to interest rate changes.
43. Slovene banks are resistant to temporary modest interest rate shocks. A two standard deviation shock to interest rates (2.4 percentage points) shows that the effective interest rate (that takes account of rigidities in balance sheets) would rise by less than 2 percentage points. As a result, interest revenue and expenses are higher in 2006 and 2007 compared to the baseline scenario, but net interest revenue falls This drop is partly offset by lower costs, such as lower provisions, reducing the negative impact on profits.16 Capital adequacy is affected more by reduced loan growth than by lower profits, which improves the capital adequacy ratio (CAR) at the time of the shock. Thus the test shows that changes on the banks’ balance sheets in response to changes in interest rates remain moderate with significant growth of loans prevailing. However, the interest rate shock used by the test may be too small biasing the results. For example, stress tests conducted by the IMF in the Financial Sector Assessment Program usually apply a shock of three standard deviations to interest rates, and the Basel Committee Amendment to the Capital Accord suggests interest rate increases between 100 and 300 basis points with stress tests.
|Profit||Return||Capital||Growth of loans||Non-bank||Growth of deposits||Non-bank||Growth of|
|in Euro mln||on Equity||Adequacy||to non-banks||loans/TA||by non-banks||deposits / TA||TA|
|Raising interest rates in the period from Q4 2006 to Q3 2007|
|Differences from the baseline scenario|
44. Foreign borrowing by banks has reduced exposure to interest rate sensitivity by lengthening maturities and reducing maturity mismatch. While about two-thirds of domestic deposits mature within one year, only 12 percent of the banks’ foreign loans are short-term. This has improved the maturity structure of banks reducing interest rate and rollover risks. Time to repricing (the remaining time until a change in benchmark rates takes effect) also shows that 94 percent of domestic deposits are repriced within one year, while at 88 percent for foreign credit the share is slightly lower. The average time to repricing of foreign assets also fell from 19 to 16 months between 2005-06, while it remained broadly stable at 6 months for foreign liabilities. As a result, foreign funding has reduced the maturity mismatch for Slovene banks, reducing sensitivity to interest rate shocks.
45. Banks can sustain moderate increases in foreign funding costs through smaller profits. A 0.5 percentage point reduction of the net interest rate margin, which corresponds to a rise in foreign funding costs by about 2 percentage points, is absorbed by banks through lower profits. The shock would leave the CAR broadly unaffected as there are no balance sheet effects assumed. However, the test may be too mild as it would leave the net interest rate margin at a level above the current EU average. An earlier IMF assessment applied a more significant shock that halved the interest margins. As a result, banks would have incurred losses, and three banks would have required a capital injection.17
|Profit (EUR mln)||Return on Equity||Capital Adequacy|
|Fall in the interest margin by 0.5 percentage points in the period from Q4 2006 to Q3 2007|
|Differences from the baseline scenario|
46. In sum, the stress test show that the Slovene banking system can easily absorb moderate interest rate and margin shocks. A sensitivity analysis for a temporary interest rate shock results in lower profits, but credit growth and capital adequacy are sustained. Foreign borrowing has increased the duration of the banks’ liabilities, therefore reducing the maturity mismatch. A jump in foreign funding costs, that reduces the net interest margin, could significantly reduce profits, but only a very dramatic fall in margins would result in losses and undercapitalization. However, the shocks are relatively mild, and larger shocks, or the combination of shocks such as a deterioration in credit quality, credit losses or a sudden stop of foreign funding can show greater vulnerabilities. This, however, remains to be tested.
Credit and currency risk
|Loans to||Loans to Nonbanks|
|Foreign claims (percent of total assets)||18.4||7.6||5.2||2.2|
|Foreign classified assets (percent of total assets)||14.4||3.0||3.6||2.1|
|Provisions (percent of foreign classified assets)||0.5||10.5||5.5||5.5|
47. Credit and currency risks arise mainly from exposures to SEE. Euro adoption and EU membership are likely to have reduced some of these risks for Slovenia, while growing exposure to SEE raises mainly credit and currency risks in a region that still needs to see a slow-down in growth.
48. The greater credit risks of foreign lending are reflected in higher provisions in Slovene banks, which should mitigate vulnerabilities. Non-resident loans are classified riskier than domestic loans by banks--provisions for loans to non-residents are about 9 percent compared to about 4 percent for all loans. This reflects the higher proportion of nonperforming loans (NPL) in some areas, such as SEE, where estimates of the NPL ratio range from 3 to 10 percent.
49. However, banks’ ability to face risks is reduced by low provisions for loans to subsidiaries of Slovene banks and to Slovene enterprises in SEE. Foreign subsidiaries originate loans in their credit portfolios and refinance them through their mother institution in Slovenia. However, the latter classify these loans as low risk.18 This may also explain why provisioning in Slovenia to non-EU countries is lower than to EU-countries. As data on onlending by these foreign subsidiaries is not available, and thus credit risk at the subsidiaries and the likelihood of a risk pass-through are hard to quantify. This can mean that credit risks abroad may be underestimated in balance sheets in Slovenia.
50. At the same time, currency risks in Slovene banks from foreign operations have been substantially reduced by euro adoption. On the funding side, euro adoption eliminated much of the currency risk, while on the lending side, indirect currency risks remain despite the fact that loans in SEE tend to be in foreign currencies.19 While the euro is the most common currency of denomination in these countries, some loans, in particular residential housing loans, are denominated in non-euro currencies, such as Swiss francs. Part of the lending may also be to unhedged borrowers, that may be vulnerable to a sudden change in market conditions or exchange rates. This in a way can substitute currency risk for credit risk in exposure to SEE. As the lending business expands rapidly during times of benign economic circumstances, the quality of this segment of lending, and the appropriate handling of a downturn, is yet to be tested
51. Overall, currency and credit risks from foreign operations of Slovene banks remain small given the still limited exposure in total assets. Despite rapid growth in lending to SEE, and potential under-provisioning for credit risks in the region, this exposure at 3 percent of the banks’ balance sheets remain small. Together with euro adoption, this suggests that overall risks from exposure to SEE are low. However, their development should be monitored closely. Loan growth to this region is high, and available information on the pass-through of credit risk from foreign lending appears limited despite well established supervisory cooperation.
52. The growing dependence of Slovenia’s bank finance on external funds increases the sector’s vulnerability to liquidity risk. Given the banks’ funding from the euro-area and their credit exposure to SEE, a sudden downturn in the latter region could have repercussions on Slovene banks. A contagion scenario, triggered for example, by a shock in some of the SEE countries, would not only deteriorate the credit quality of Slovene bank assets, but could at the same time cause a sudden stop of foreign funding. To assess these risks, the following traces the origins of capital flows to Slovenia and evaluates a potential impact of a withdrawal of foreign lending on the banks’ liquidity ratio and profitability.
53. The vulnerability to contagion in Slovenia and in SEE is increased by concentration of foreign funding to banks in a few neighboring EU countries. Austria is the most dominant provider of capital to Slovenia, as to SEE in general, followed by Germany. Most of the investments are bank loans reflecting the relatively underdeveloped financial markets. The dominance of a few countries in the bank loans may reflect the distribution of foreign equity participations in Slovenia--most of the 35 percent of foreign held-equity originates in Austria and Germany. As the same banks are active in other countries in the region as well, potential for contagion risks is increased. By the end of 2005, half of all Eastern European foreign owned bank assets were concentrated in eight bank groups.
Figure 1.Foreign Bank and Portfolio Investment by Residence of Investor, 2005
Sources: BIS; CPIS; and staff estimates.
1/ Albania, Bosnia and Herzegovina, Croatia, Macedonia, Serbia and Montenegro. Other euro-area includes Netherlands, Portugal, Spain; Non-euro Europe includes Denmark, Norway, Sweden, Switzerland, United Kingdom. Data unavailable or confidential for bank loans from Norway and portfolio investments from United Kingdom.
Eastern European Bank Asset Structure, by Key Bank Groups, end-2005 1/
Source: RZB Group (2006).
1/ Markets include Albania, Belarus, Bosnia and Herzegovina, Bulgaria, Croatia, Czech Republic, Estonia, Hungary, Kosovo, Latvia, Lithuania, Poland, Romania, Russia, Serbia (data presented are for Serbia and Montenegro), Slovakia, Slovenia, and Ukraine.
54. The vulnerability of Slovene banks to sudden stops was tested with changes to the liquidity ratio. This stress test assesses the change in the liquidity ratio (which is defined as the ratio between liquid assets and short-term liabilities) and pre-tax profits in two scenarios--a withdrawal of liabilities to foreign banks by 20 and 100 percent. The reduction in liabilities to foreign banks is followed by an adjustment on the asset side according to their degree of liquidity. Depending on the liquidity need and the available liquid assets, the stress test shows whether banks need to diminish their credit activity. However, this test can slightly underestimate the impact of sudden stops, as this may also have repercussions on the whole economy.
|Actual||Foreign liquidity shock of 20%||Foreign liquidity shock of 100%|
|Total||Total||Domestic Banks||Foreign||Total||Domestic Banks||Foreign||Total||Domestic Banks||Foreign|
|(In Euro bln)||Dec-05||(Sep 2006)||Large||Small||Banks||Large||Small||Banks||Large||Small||Banks|
|Securities held for trading||2.0||1.4||1.0||0.2||0.2||1.4||1.0||0.2||0.2||1.1||0.2||0.2|
|Short-term NBS assets||6.4||6.4||4.3||0.9||1.2||6.4||4.3||0.9||1.2||5.6||4.3||0.4||0.4|
|Liabilities to foreign banks||9.8||9.2||4.9||0.2||4.1||7.3||3.9||0.1||3.3||−||−||−||−|
|Liabilities to domestic ban||0.6||0.7||0.3||0.3||0.7||0.7||0.3||0.3||0.7||0.3||0.3||0.3||0.1|
|ST Liabilities to NBS||9.5||8.2||5.8||1.1||1.3||8.2||5.8||1.1||1.3||8.2||5.8||1.1||1.3|
|Liabilities to foreign banks,||0.3||0.3||0.2||0.1||0.5||0.2||0.2||0.0||0.5||−||−||−||−|
55. Foreign-owned banks are most vulnerable to a withdrawal of foreign funding. About one half of total assets and liabilities in the Slovene banking system are short-term. The share of short-term liabilities provided from abroad is 40 percent for domestic banks, and three quarters for foreign-owned banks. As a result, smaller banks have the most comfortable liquidity ratio, while it is lowest in the foreign-owned banks. This may reflect the foreign-owned banks’ role of transferring funds from their parent banks to long-term loans in Slovenia. The stress test, that assumes that short-term assets can be liquidated without any problems, shows that domestic banks can sustain the complete withdrawal of foreign funding without incurring liquidity problems. In contrast, foreign owned banks, that depend on foreign funding, are more vulnerable. In case of a complete withdrawal, foreign banks would suffer from an insufficient liquidity ratio, and would need to liquidate their long-term assets. However, the stress-test scenario assumes that foreign subsidiaries are abandoned by their mother institutions, which is unlikely.
|Interest Spread||Before Shock||After Shock||In Profit|
|(In percent)||(EUR min)||(EUR min)||(EUR min)|
|Foreign liquidity shock of 20 percent|
|Large domestic banks||4.25||238||196||−42|
|Small domestic banks||2.71||38||38||0|
|Foreign liquidity shock of 100 percent|
|Large domestic banks||4.25||238||29||−209|
|Small domestic banks||2.71||38||33||−4|
56. Foreign banks are most vulnerable to sudden stops, with the larger shock causing a profound loss of income. With 20 percent withdrawal of liabilities to foreign banks, six of the nine foreign-owned banks would make losses. In case of a complete withdrawal, all foreign owned banks and two large domestic banks would incur losses.
57. If feedback effects are incorporated, a liquidity shock would also impair large domestic banks. The current test assumes that banks can choose liquidating interbank loans, while the effect on the counterparty’s balance sheet is not taken into account. In case of a full withdrawal of foreign funding, the complete liquidation of all interbank assets (as suggested in the table) might prove unrealistic as this would send shock waves throughout the banking system. Taking feedback effects of liquidating all interbank loans into account, large domestic banks would be left with insufficient liquidity ratios.
58. In sum, the analysis shows that liquidity risk in Slovenia is low. Although the tests point to some vulnerability to sudden stop shocks, their occurrence in Slovenia is likely to be small. While a shock in SEE may cause foreign banks to reduce exposure in the region, they are likely to differentiate with Slovenia given its membership in the euro area and the EU. Its exposure in SEE is also still modest to create major concerns for the banking system in Slovenia. However, this is not to say that Slovenia would not be impacted by a slow-down in foreign credit given its high reliance on foreign borrowing from a few sources to finance asset growth. Thus diversification of the funding base would spread the risks better and reduce the vulnerability to liquidity risk.
59. While vulnerabilities to various risks from foreign exposure seem contained in Slovenia, the rise in cross-border activities merits close monitoring. While increasing the dependence on foreign funds, borrowing from abroad has diversified the funding base and contributed to a lower maturity mismatch reducing exposure to interest rate risks. The strong expansion of lending abroad has mitigated the margin pressure on Slovene banks, but it comes at the expense of higher credit risks. Although this exposure remains small in total, the current provisions may not fully cover these risks. A deterioration of foreign credit quality, possibly upon the end of the current credit boom in SEE, could cut into profits. Despite a concentration of funding from a few sources in Europe exposed to elsewhere in Eastern Europe, and large impact of sudden stops with stress tests, contagion risks remain small given Slovenia’s euro area membership.
60. Closer monitoring, in particular with regard to credit risks, and broader stress tests could add to the understanding of risks from cross-border lending and borrowing. The analysis on credit risk is limited by data availabilities, and, more generally by perhaps benign assumptions. Better reporting of credit risks in SEE would give more information on country exposure and currency composition from foreign lending and borrowing. More thorough information about the credit quality of non-resident loans and the loan portfolios of Slovene banks’ foreign subsidiaries are also necessary to monitor the appropriateness of provisions.20 Stress tests could be extended to analyze a scenario which combines a sudden decline in foreign funds with a simultaneous deterioration of credit quality of foreign loans, including those to foreign subsidiaries of Slovene banks.
61. Cross-border supervisory cooperation will also remain important in detecting vulnerabilities. Supervisory coordination between EU countries could serve as benchmark to establish a level playing field of foreign banks active in non-EU countries.21 Besides cross-border interagency communication, memoranda of understanding could also encompass crisis prevention.22
Interest rate stress test
The interest rate shock used--a temporary interest rate increase of 2.4 percentage points--corresponds to two standard deviations of the variance in the last 11.5 years. This scenario assumes that the two main reference interest rates, the yield on 60-day tolar bills and the one-year EURIBOR, are raised for four consecutive quarters (Q4 2006-Q3 2007) before receding to their previous level. Upon the shock, the income implications are derived based on the duration of assets and liabilities. Other input variables are taken from the Economic Projections based on a structural model of the Bank of Slovenia and published in the Monetary Policy Report, October 2006.
The setup of the stress test implies that the interest rate shock reduces the demand for loans and stimulates savings, albeit with a lag and delayed pass-through of higher interest rates on deposits. On the one hand, the non-banking sector responds to the increase in interest rates by reducing new borrowing. The rate of growth of loans to the non-banking sector in 2006 is 0.4 percentage points lower than in the baseline scenario. The growth of total assets follows the growth of loans to the non-banking sector. Current loans require longer to adjust to the new circumstances, which is why most of the impact of the higher interest rates manifests itself only in 2007. On the other hand, higher interest rates stimulate saving, although with a delay. In the case of a higher rise in interest rates, the non-banking sector would allocate disposable income for the early repayment of current loans rather than for deposits. Higher interest rates could encourage the transfer of funds from alternative investments into bank deposits, although this shift is not immediate and only partial.
In another test, a net interest rate margin shock of 0.5 percentage point is assumed to take place in the fourth quarter of 2006 and reverse one year later. As opposed to the interest rate stress test, the interest margin shock is modeled to affect only interest expenses, while the interest rate shock affects both interest expenses and revenue. The margin shock is assumed to leave the balance sheet structure unaffected. The drop in the interest rate margin tries to mimick a rise in the cost of foreign funding. A corresponding rise in foreign interest rates would be highest in the first quarter of the shock, rising by 2.3 percentage points. In subsequent quarters, the interest rate rises less (2 percentage points in the second quarter, 1.9 in the third and 1.8 in the fourth quarter) due to base effects.
Liquidity stress test
The liquidity ratio is defined as the ratio between liquid assets and short-term liabilities. Based on Boss and others (2004), liquid assets are defined as the sum of interbank assets, cash reserves, government bonds available for sale, debt instruments and equities held for trading, and other short-term non-banking sector assets. Short-term liabilities are defined as the sum of liabilities to foreign banks, liabilities to domestic banks and short-term liabilities to the non-banking sector.23
The liquidity shock assumes that foreign banks withdraw their short-term funding and force Slovene banks to liquidate their short-term assets accordingly. Upon the reduction of short-term liabilities to foreign banks, the assets side is adjusted by liquidating the most liquid assets first before rationing long-term credit business. Depending on the need to reduce short- and long-term credit activities, the impact on profits is calculated on base of current weighted interest margins.
Bank of Slovenia2006aFinancial Stability Report,May2006.
Bank of Slovenia2006bMonetary Policy Report,October2006.
BelaischA.KodresL.LevyJ. and UbideA.2001“Euro-Area Banking at the Crossroads,”IMF Working Paper No. 01/28.(Washington:International Monetary Fund).
BossM. G. Krenn M. SchwaigerM. and W.Wegschaider2004“Stress Testing the Austrian Banking System,”Bank-Archiv Austrian Society for Bank-Research ÖBA 11/2004pp 841–56.
International Monetary Fund2004FSSA Update,IMF Country Report No. 04/137(Washington:International Monetary Fund)
Prepared by Jochen Andritzky (MCM) in collaboration with Tomaz Kosak, Financial Stability Department, Bank of Slovenia.
The adoption of the International Financial Reporting Standards (FRS) in 2006 is expected to support growth in profits in the medium term.
See IMF (2004), p. 14.
Loans to subsidiaries are classified as A-rated on a rating scale from A to E.
78 percent of lending in Croatia, and 85 percent in Serbia, (in particular so for long-term lending) is denominated in non-domestic currency.
The 2004 FSAP recommendations have already pointed out that the banking supervision should strengthen its risk focus and ensure that pricing of risk and provisioning are appropriate. Furthermore, supervisors should give consideration to how prudential tools, including provisioning policies, could be used to address risks.
See IMF (2007).
See Boss and others (2004).