Democratic Republic of S�o Tom� and Pr�ncipe: Selected Issues

Abstract

Democratic Republic of S�o Tom� and Pr�ncipe: Selected Issues

Mainstreaming Macro-Financial Linkages1

A. The Landscape of the Financial System

1. São Tomé and Príncipe has a small financial system with money supply (M2) around US$90 million at the end of 2015.2 Similar to other small financial systems, concentration of assets and operating costs are high, coupled with relatively low asset quality and lower profitability. However, recent trends show further deterioration in key areas—increasing NPLs and large and growing excess liquidity—from averages in other small financial systems (section IV). Private sector credit (as a percent of GDP) is low yet comparable to other small states, including sub-Saharan African (SSA) small states, at 28.3 percent in 2014. Commercial banks account for approximately 98 percent of financial sector assets, and a large majority of these banks (6 out of 7) are foreign-owned; in addition, there are two insurance companies and four small consumer lenders operating in São Tomé and Príncipe. The banking sector is highly concentrated, with the largest bank holding almost 50 percent of total assets, and the largest three banks holding almost 75 percent of total assets. The analysis in this paper will, therefore, focus on the banking sector.

2. The size of the banking sector expanded rapidly following its liberalization in the early 2000s. The Banco Internacional de São Tomé e Príncipe (BISTP) was the only commercial bank since 1993 until three new entrants in 2004 after the liberalization. The number of banks increased to eight by 2008, in anticipation of increased economic activity in relation to commercial oil production. With the uncertainty of commercial oil production, the central bank of São Tomé and Príncipe (BCSTP) has had to put three banks under administration within a decade of granting their license, underscoring weak regulatory and supervisory environment that could not cope with the rapid increase in the number of banks.

3. There has however, been notable progress in banking supervision since 2012. The BCSTP began a full cycle of on-site inspections in 2012, which was completed in 2015; a new cycle of inspections commenced in 2016. Improvements in BCSTP’s banking supervision framework have helped identify banks with inadequate capital, obtain more reliable and realistic estimates of NPLs and provisions. Both the use of supervisory instruments and the monitoring of banks’ adherence to the prudential standards have also improved.

B. Financial Sector Stability Analysis

4. The two main concerns from a financial sector stability standpoint are excess liquidity and deterioration in asset quality. São Toméan banks started to experience excess liquidity, increasingly over the last three years. Consolidated banks’ balance sheet (Table 1) illustrates the feeble growth in loans as well as a substantial increase in liquid assets in cash and treasury bills. The slow credit growth is indicative of banks’ risk-averse stance, particularly, given an increase in non-performing loans and uncertainties surrounding delays in addressing collateral enforcement and default risks through the judiciary system. In 2015, the ratio of liquid assets to total assets reached 52 percent. Despite the stagnation in credit growth, deposits have been increasing steadily and grew by more than 20 percent in 2015. Of total deposits, almost 70 percent are in demand deposits (versus time deposits) and around 40 percent are in foreign currency.

Table 1.

Key Developments in Banks Balance Sheet

article image
Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

5. A preliminary asset quality review shows that the quality of the banks’ aggregated loan book deteriorated significantly since the last Article IV in 2013 (Figure 2). The ratio of NPLs to total loans almost doubled from around 16 percent to almost 30 percent, largely driven by substandard and loss loans, which indicates further deterioration in asset quality. However, provisions more than doubled and there is heterogeneity among banks. That notwithstanding, bank-by-bank NPLs show a worrying picture, with only one bank with an NPL ratio of less than 10 percent. Two banks have NPLs close to 50 percent while the remaining banks have NPLs ranging between 20 and 30 percent, reflecting similar trends in other small financial systems (Figure 3).

Figure 1.
Figure 1.

Key Financial Sector Indicators

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.
Figure 2.
Figure 2.

Asset Quality of Banks

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.
Figure 3.
Figure 3.

Dispersion of NPLs/Total Loans by Banks (2015)

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: Säo Tomé and Príncipe authorities data; and IMF staff estimates.

6. Other sector-wide prudential indicators have shown some improvement following supervisory action. The capital to risk-weighted assets ratio increased from 20.3 percent in 2012 to 24.1 percent in 2015 as three banks were asked by BCSTP to inject additional capital after supervision activity identified that their capital bases were inadequate, relative to the prudential ratios. Consequently, all active banks have met or exceeded the regulatory threshold of 12 percent. Nonetheless, the capital to assets ratio continued to decline and stood at 15.5 percent at the end of 2015.

7. Compared with other very small financial systems, the interest rate spread has been higher in São Tomé and Príncipe, hovering around 12–13 percent versus the 9.25 percent average in small financial systems (SIP Section IV). The margin decreased from 18 percent in 2010 to 13 percent in 2015. While declining since 2010, the lending rate on average was around 17-18 percent and deposit rate around 5-6 percent in 2015. The high lending rate, despite the peg, reflect both the high operating costs and increase in banks’ risk aversion as a result of the recent increase in NPLs.

8. Dollarization has halved in the last 5 years (Figure 5). Since the official pegging of dobra to the euro, the share of loans denominated in local currency increased while those in foreign currency decreased. As a percentage of the total portfolio, deposits and loans in foreign currency decreased from levels above 60 percent in 2010 to around 30 percent in 2015, with the downward trend becoming more visible after 2013.

Figure 4.
Figure 4.

Trends in Interest Rates

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.
Figure 5.
Figure 5.

Total Credit

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

9. The sectoral composition of banks’ consolidated loan portfolio remained almost unchanged over 2013–15, while the loans-to-GDP ratio decreased from 34 to 27 percent during the same time period (Figure 6). Loans for construction, trade and consumption account for the largest proportions of credit by economic sector. While the credit growth in the past decade (until 2013) could be attributed to prospects of oil production, sectoral allocation of credit has not changed since credit growth started slowing down in 2013.

Figure 6.
Figure 6.

Loans by Sector

(Percent of Total Loans)

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

10. At the individual bank level, credit exposure to sectors and the size of the credit portfolio vary greatly (Figure 6). For example, in 2015, credit to the construction sector was provided mainly by one bank, and to a large degree the same is true for lending to tourism and manufacturing; this credit concentration by sector at the individual bank level increase banks’ vulnerability to sector-specific shocks.

11. Low profitability of São Toméan banks has been mostly driven by noninterest expenses and provisions for loan losses, which have both increased in 2015 (Table 2). One of the problems faced by banks in São Tomé and Príncipe is high operating costs, largely attributable to utilities (particularly electricity) and the technology infrastructure (such as information systems and the ATM-related costs).

Table 2.

Breakdown of Bank Profitability

article image
Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

12. Analysis of São Tomé and Príncipe’s financial soundness indicators also indicates problems in earnings. Most banks have reported negative earnings since 2013 and these losses have been rising (Table 2 and Figure 7). The expense to income ratio has also been increasing. The increase in provisions does not match growth in NPLs, fluctuating at around 100 percent of NPLs in 2015 compared to above 150 percent in 2013. While this could be partially explained by the increase in the NPLs, it flags a need to increase provisioning for problem loans by banks.

Figure 7.
Figure 7.

Expenses vs. Profitability (2011–15)

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

C. Macro Financial Linkages

13. Linkages between developments in the financial sector and the overall economy are bi-directional. In the context of examining these bi-directional linkages, we focus on the real, fiscal, external and monetary sectors. Using the banking sector data, bank-by-bank balance sheet analysis as well as discussions with market participants, these linkages and their transmission channels (including spillovers from the financial sector to the rest of the economy and spillbacks from the rest of the economy to the financial sector) were identified (Figure 8).

Figure 8.
Figure 8.

Macro-Financial Linkages in São Tomé and Príncipe

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

14. The real and financial sector linkages are primarily through the credit channel. A breakdown of the loan portfolio in 2015 show that trade (24 percent), construction (24 percent) and consumption (21 percent) received around 70 percent of the credit to the economy (Figure 9), and this pattern has not changed much in the last two years. On the other hand, credit to the agricultural and tourism sectors remained low at about 1 percent each. Expansions in the agriculture and tourism sectors are largely externally-financed by grants and FDI respectively. The share of credit to the public sector, including the local government and public non-financial corporations, is miniscule—1 percent of gross claims in 2015 (Figure 10). NPLs in the top three domestic credit-receiving sectors are also relatively higher than in other sectors; in 2013, NPLs in the trade sector were disproportionately higher (more than 50 percent) but, in 2015, there were also significant increases in NPLs in consumption and construction.

Figure 9.
Figure 9.

Loan and NPLs by Sector

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.
Figure 10.
Figure 10.

Loan Distribution by Type of Borrower

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

15. The banking system has minimum direct exposure to the public sector, while the private sector has been the largest borrower in the last three years, followed by individuals/households (Figure 10). Credit to public sector has been at very low levels in 2013 and 2014, with an increase in 2015 due to available credit lines to EMEA). While credit to households grew rapidly in 2010/11, it has been contracting since 2013. High levels of NPLs to households reflect both expansions in credit in anticipation of increased incomes that normally come with commercial oil production, as well as lack of bankable investment opportunities for the banks. Poor lending standards and credit stance and in adequate supervisions during the credit boom (up until 2013) also contributed to the rapid accumulation of NPLs.

16. There are also significant spillbacks from the real sector to the financial sector. For example, inefficiencies in energy production has contributed significantly to high operating costs for most financial institutions which are charged higher tariffs in the tariff structure. At the end of 2015, on average, electricity costs constituted about 40 percent of total operating cost of most financial institutions and this has been the trend in the last decade. These high operating costs in part explain the low profitability and high interest rate spreads in the banking sector.

17. Some of the locomotive sectors of economic growth in the economy benefit less than optimally from bank credit and access to finance. The construction sector with assets that are used as collateral was the recipient of almost a quarter of total bank credit, even though its contribution to real GDP was 7 percent in 2015. Conversely, credit to the agriculture sector made up the smallest share (1 percent) even though the sector—including fisheries—contributes close to 15 percent to total GDP, ranking as the second largest sector, and is a key employer. Lending for wholesale and retail trading averages roughly a quarter of total outstanding credit to the private sector, consistent with the distributive trades’ dominance in the economy (contributing about a quarter to real GDP). Similarly, tourism received around 1 percent of domestic credit, in line with its contribution to GDP (around 2 percent), underscoring its untapped potential.

18. One important factor that restrains the real-financial sector linkages is the high degree of informality in São Tomé and Príncipe’s economy. Informality limits access of small and medium-sized enterprises to finance as well as financial inclusion. However, the establishment of a credit registry partially decreased the information asymmetries and associated costs in the formal sector.

19. The linkages between the fiscal and financial sectors are relatively weak (Figure 11). While the government does not directly rely on domestic financing, the main linkages with the fiscal sector come from accumulation of large domestic arrears to key businesses and potential costs to the budget from restructuring/recapitalization of banks. A less direct linkage comes through private companies which provide services to the government, particularly financing for investment projects. On the other hand, where the private sector mostly depends on government spending on goods and services, any delays in budget implementation and any accumulation of arrears to suppliers by the government could impact on the ability of these government contractors to service loans. Stress tests under adverse-shock scenarios show that restructuring/recapitalization costs to the budget could be in the range of 1½ to 3 percent of GDP. Performance in the banking sector has a direct impact on government revenues, in particular, from corporate taxes. Recent low profitability of banks has affected government domestic revenue mobilization negatively. Only two banks paid corporate taxes in 2013 and this drop to just a bank paying taxes in 2014 and 2015.

Figure 11.
Figure 11.

Simulated Financial Sector Impact on the Budget

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

20. The complications that arise from excess liquidity in the banking sector highlight the bi-directional linkages between the monetary and financial sectors. Excess liquidity of financial institutions weakens monetary policy by undermining the effectiveness of tools available to the central bank. For instance, any BCSTP decision to change the reserve requirement would only have very limited impact on credit. Similarly, the monetary transmission mechanism is also weakened, limiting the impact of changes to interest rates.

21. Large presence of foreign-owned banks could potentially pose risks to reserves and the currency peg arrangement. In a banking sector where around 50 percent of total assets are held by foreign-owned banks, the central bank faces the risk that these banks may decide to move their investments abroad given scarce domestic investment opportunities and low profitability. Such a situation would lead to increases in capital outflows and demand for selling local currency, and hence induce pressure on the currency and the peg. Similarly, any reversal of the downward trend in dollarization upturn in foreign currency denominated loans would affect international reserves and induce stress on the current exchange rate peg arrangement.

22. The development of domestic debt market and introduction of new instruments could strengthen the monetary-financial sector linkages. The BCSTP recently strengthened its efforts to introduce domestic debt instruments through issuances of two Dobra-denominated treasury bills in 2015 and 2016. These are expected to be instrumental in improving the interbank market, which has not been used by the banks so far due to high risk aversion and the stigma associated with accessing the interbank market. The issuance of treasury bills and bonds are expected to serve as collateral to back lending in the interbank market. Additionally, development of the treasury bills market—developing a secondary market and the number of bills on offer with different maturities—would help the BCSTP build a yield curve, which would decrease both uncertainty in the market and the risk premium in investments. This would also strengthen the BCSTP’s liquidity management tools and options such as open market operations in the medium and long term. However, while developing the debt market, there should also be a clear borrowing strategy by the government and issuance should be aligned with fiscal policy and debt sustainability.

23. There are a number of bi-directional linkages between the external and financial sector, primarily through exports and trade. Strengthening of the external and financial sector linkages could play a key role in promoting growth through exports, which would also have positive implications for the rest of the economy.

24. The tourism sector is expected to play an important role as a key pillar of economic growth in the medium term. The financial sector could support tourism by providing credit to and investing in the sector and improving the sector’s competitiveness by offering a range of financial services to tourists. In São Tomé and Príncipe, both of these support mechanisms are weak; credit to the tourism sector is marginal and financial services available to tourists and service-providers are quite limited. Financial sector innovation would help improve São Tomé and Príncipe’s competitiveness.

25. The financial sector could also play a critical role in facilitating trade, particularly with respect to payments. Currently, foreign exchange is limited and not always available to banks, weakening their capacity to support importers. This could partially be explained by the BCSTP’s strict control to minimize open positions. Financial instruments to support trade are also very limited and as a result importers and exporters face practical difficulties. For instance, letters of credit are not used for payments, and instead businesses resort to the high-risk option of transferring money in advance for imports and having no control over the trade.

D. Macro Stress Tests

26. To identify areas of potential vulnerability in the banking sector that would become evident under adverse scenarios, macro stress tests were conducted as part of the macro-financial linkages study in São Tomé and Príncipe. These tests utilize a set of statistical techniques to help assess the vulnerability of financial institutions and financial systems to exceptional but plausible events. The objective is to make risks more transparent by estimating potential losses in abnormal circumstances. At the moment, these stress tests are not part of the BCSTP’s supervisory framework, although the authorities are familiarizing themselves and experimenting with these tests to complement surveillance of the financial sector.

27. The exercise draws on the Cihak (2007) framework, adjusted for São Tomé and Príncipe banking sector’s specificities and data. Four types of risks are analyzed: credit risk, liquidity risk, market risk (interest and foreign exchange) and income risk. The impact of minor, moderate and major shocks on São Toméan banks is tested for each risk both at the aggregate and individual bank level. The tests are based on end-2015 banking sector data provided by the BCSTP.

28. Stress tests suggest that in São Tomé and Príncipe, credit risk is the most relevant given the limited debt instruments, excess liquidity and currency peg. Vulnerabilities come from of the accumulation of NPLs and high concentration of credit exposure. The results are discussed in the remainder of this section.

29. The banking sector’s resilience to solvency risk was assessed by stimulating the impact of an overall deterioration in loan quality on capital adequacy ratio (CAR) per bank and for the whole sector. A minor shock of 20 percent, a moderate shock of 40 percent and a major shock of 70 percent increase in total NPLs were applied and a provisioning rate of 100 percent for the new NPLs was assumed. The results show that the banking system in aggregate would remain resilient to minor, moderate and major shocks, with CAR staying above the 12 percent minimum requirement CAR (Figure 12). However, at the individual bank level, the major shock leaves two foreign banks with CARs below the minimum threshold. In an extreme case scenario where NPLs are assumed to increase by 90 percent, the banking sector’s CAR drops below the minimum requirement, driven by deterioration in foreign banks’ balance sheets.

Figure 12.
Figure 12.

Effects of Shocks to NPLs on Capital Adequacy

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

30. The sectoral shocks to NPLs aim to better assess the vulnerability faced by banks, using recent trends in NPLs by sector (Table 3 and Figure 13). The assumed increases in NPLs were based on the observed historical growth of NPLs by sector since 2012. Under the minor shock scenario, two banks’ CARs decrease to around 11 percent. When NPLs increase under the major shock scenario, the CAR at the aggregate level deteriorates, with one bank becoming insolvent due to its large exposure to the construction sector and four of the foreign banks remain solvent with CARs above 12 percent.

Table 3.

Sectoral Impact of Shocks to NPL

article image
Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.
Figure 13.
Figure 13.

Credit Risk: Impact of Sectoral Increases in NPLs on Capital Adequacy

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

31. Large exposure to few borrowers exposes banks to concentration risk as a default of large borrowers would lead to CAR dropping below 12 percent. The results of the stress tests show that the entire banking sector is exposed to the concentration risk (Figure 14). The overall CAR ratio would drop below the minimum capital requirement after the default of the two or three largest borrowers of each bank. In the case of two largest borrowers defaulting, two foreign banks experience substantial difficulties in capital adequacy, whereas with the default of the three largest borrowers, overall CAR ratio drops to 10 percent.

Figure 14.
Figure 14.

Concentration Risk

Citation: IMF Staff Country Reports 2016, 175; 10.5089/9781498308373.002.A002

Sources: São Tomé and Príncipe authorities’ data; and IMF staff estimates.

32. For assessing the liquidity risk, simple liquidity tests are more suitable for São Tomé and Príncipe given that the financial sector is in its early stages of development. Moreover, the interbank market is almost non-existent. Therefore, the risk of drying-up of liquidity in interbank operations is negligible. Simple liquidity tests model a simple liquidity drain that affects all banks in the system proportionally. In a financial system with excess liquidity like São Tomé and Príncipe, the results show that liquidity risk is low for the banks. In case of a liquidity run, all banks would not face a liquidity problem in the first 5 days or the need to access external financing for liquidity.

33. Given the shallowness of government debt market and its short history, São Tomé and Príncipe banks are not sensitive to changes in bond interest rates and do not face any significant interest rate risk. Similarly, the current foreign exchange risk is quite limited in the presence of a currency peg. While there is still dollarization in the system, it is at much lower levels (30 percent). Thus, not presenting a significant risk to the banks through exchange rate fluctuations. Neither direct nor indirect foreign exchange risk results in banks’ CARs dropping to below the minimum requirement of 12 percent.

E. Policy Implications

34. Continued vigilance over the financial sector is warranted given the adverse impact of macro-related risks on banks and the negative feedback loop to the overall economy. Ensuring the health of the financial sector would bolster its role in supporting economic growth. In light of the recent developments regarding high NPLs, low profitability and low financial intermediation, the banking sector would benefit from a credible NPL resolution strategy. It is also necessary to address structural bottlenecks in the economy such as lengthy court resolution processes, lack of collateral, large informal sector and low financial literacy levels.

35. Deterioration in asset quality and NPLs are immediate risks to financial sector stability and soundness. The BCSTP’s prospective NPL resolution strategy and increased supervision will be keys to minimizing spillovers from the banking sector to the rest of the economy. Enhancing both on and off-site supervision and tightening and enforcing prudential norms are critical to containing credit risk and avoiding liquidation costs for the monetary authority and fiscal sector. In the presence of banking sector vulnerabilities, there is also the need to consider the introduction of emergency lending assistance (ELA) arrangements.

36. The financial sector could play an instrumental role in unlocking São Tomé and Príncipe’s growth potential. Currently, the relative importance of the various sectors’ contributions to the economy (in terms of GDP and employment) is not fully aligned with the credit distribution trends. Therefore, while high NPLs are reasonably expected to restrain credit expansion, once the banking sector recovers and becomes more stable, domestic credit should support the country’s growth strategy. In this respect, supporting sectors like agriculture and tourism would benefit both the external and the real sectors.

37. Macro stress tests show that credit risk is currently the biggest threat to financial sector stability. Simulated shocks lead to rapid accumulation of additional NPLs, at the aggregate and individual bank levels and concentration risk. In terms of sectoral allocation, banks’ portfolios vary notably, such that certain shocks affect the banks differently. Bank-level results of stress tests therefore shed light on some risks that disappear at the aggregate level. Consequently, conducting stress tests regularly to complement on and off-site supervision would help ensure the soundness and stability of the sector.

1

This note was prepared by Burcu Hacibedel (FIN).

2

See section IV for detailed discussions of small financial systems.

Democratic Republic of São Tomé and Príncipe: Selected Issues
Author: International Monetary Fund. African Dept.