Resilience and Growth in the Small States of the Pacific
Chapter

Chapter 17. Determinants of Interest Rate Spreads in the Solomon Islands

Author(s):
Hoe Khor, Roger Kronenberg, and Patrizia Tumbarello
Published Date:
August 2016
Share
  • ShareShare
Show Summary Details
Author(s)
Nooman Rebei

Deposit-lending rate spreads are large in the Solomon Islands compared with the rest of the Pacific island countries, despite their recent decline. Several studies have looked at the causes and implications of high spreads in other regions. However, few have addressed the banking sector in the Pacific islands, and almost none have considered the case of the Solomon Islands. This chapter, which examines the determinants of interest rate spreads in the Solomon Islands, aims to address this gap. Interest rate spreads are very heterogeneous in the region, and finding common lessons from a panel analysis can be misleading (Figure 17.1). Unlike previous empirical studies, this chapter focuses on local banks by using detailed quarterly bank and macroeconomic data.

Figure 17.1Interest Rate Spreads of Commercial Banks, 2014

(Annualized spreads)

Sources: Country authorities; IMF, Monetary and Financial Statistics and International Financial Statistics databases; and IMF staff calculations.

Note: Annualized interest rates. For Kiribati and Marshall Islands data are for 2011.

High bank spreads, among other factors, hinder the private sector’s access to credit, which is an impediment to inclusive growth (Figure 17.2). It is therefore important to identify the sources of high spreads in the Solomon Islands to formulate potential reforms that can be implemented.

Figure 17.2Financial Development and Inequality, Average 1990–2013

Sources: Asian Development Bank, Key Indicators for Asia and the Pacific; IMF, Financial Access Survey; World Bank, World Development Indicators; and IMF staff calculations.

1 A value of 1 indicates maximum inequality.

This chapter begins by providing the background and theory on interest rate spreads and their determinants. It then describes the empirical methodology and discusses the estimation results. This is followed by an examination of the structure of the banking sector in the Solomon Islands and an investigation into potential collusion in the sector. Finally, policy recommendations are offered.

Factors Behind High Interest Rate Spreads

To capture the different aspects of the banking sector in the Solomon Islands, bank-level data collected and published on a quarterly basis by the Central Bank of Solomon Islands (CBSI) are used. Our panel data set contains quarterly balance sheet and income statement data on every commercial bank from the first quarter of 2009 to the third quarter of 2013. Data from the IMF’s World Economic Outlook and International Financial Statistics databases were used to identify other macro-economic aggregates.

Bank spreads measure the gap between the amounts a bank pays the providers of funds and what it receives from users of bank credit. The literature distinguishes among different definitions of bank spreads. The most common are a narrow definition, which is described as the difference between interest income over loans and the interest expense over deposits; and a broad definition, which corresponds to a bank’s total interest income minus total interest expense, divided by total interest-bearing assets.1

Figures 17.3 and 17.4 show that from the first quarter of 2009 to the third quarter of 2013 bank-specific interest rate spreads in the Solomon Islands remained high regardless of the adopted methodology to measure them. Figure 17.3 shows some persistence of bank spreads, which may indicate low competition in the banking industry. This raises concerns about the effectiveness of the credit channel for monetary policy transmission and would influence the appropriate policy stance for the monetary authorities.2 It may also raise concerns about financial stability because investors tend to undertake risky projects to compensate for high lending rates, leading to increased default risks. This is the “concentration-fragility” view of the relationship between market power and bank soundness, which holds that a concentrated market weakens stability. Another view—the “concentration-stability” view—argues that larger banks in concentrated banking sectors reduce financial fragility because, among other reasons, they tend to increase profits and build up high capital buffers. This makes them less prone to liquidity and macroeconomic shocks (see, for example, Berger 1995; Uhde and Heimeshoff 2009; and Mirzaei, Moore, and Liu 2013).

Figure 17.3Commercial Bank Spreads, Narrow Definition

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

Note: The interest rate spreads for the three banks, Australia and New Zealand Banking Group Limited (ANZ), Bank South Pacific (BSP), and Westpac Banking Corporation (WBC), are calculated by taking the total interest received by banks on loans during one quarter divided by the total loans for that period and subtracting from the result the total interest paid on deposits throughout the quarter divided by total deposits. Numbers are displayed as quarterly percentages.

Figure 17.4Commercial Bank Spreads, Broad Definition

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

Note: The interest rate spreads for the three banks in the graph are defined as the difference between the quarterly interest income and quarterly interest expense divided by the total amount of assets during the same period.

The key determinants of interest rate spreads of banks in the Solomon Islands are now examined using a set of bank-specific variables, the main banking industry characteristics, and macroeconomic conditions. Different authors use the same classification of variable categories, including Gelos (2006) for Latin America, Crowley (2007) for the English-speaking African countries, and Samuel and Valderrama (2006) for Caribbean countries. The selection of variables was guided by the abundant existing literature, but data availability for banks in the Solomon Islands is taken into consideration.

Characteristics of Banks in the Solomon Islands and the South Pacific Region

Before focusing on bank-specific indicators, we start with comparing selected indicators of banks in the Solomon Islands with those in the region (Chapter 2 provides additional cross-country comparisons). Although, this is not reflected in the empirical analysis—where we consider only time series for the domestic banks—this should shed some light on the general characteristics of banks in the Solomon Islands and the major variables that should be identified as potential sources of high interest rate spreads.

These banks show clear evidence of high profitability compared with regional and international standards. They also show a higher degree of risk aversion owing to the excessive accumulated levels of equity to average assets. Credit risk, as measured by loan loss to gross loans, is moderate compared with the South Pacific region in recent years. Moreover, banks in the Solomon Islands exhibit relatively prominent salary and other noninterest expenses to average assets compared with other island countries in the South Pacific (Table 17.1). More disaggregated quarterly bank-specific data are used in the following sections to better understand the determinants of bank spreads in the Solomon Islands.

Table 17.1Selected Bank Indicators
Number of Commercial BanksReturn on Average AssetsRatio of Equity to Average AssetsRatio of Loan Loss Reserves to Gross LoansRatio of Gross Noninterest Expense to Average AssetsRatio of Nonsalary, Noninterest Expense to Average Assets
Samoa45.418.33.95.63.2
Solomon Islands37.017.83.56.33.4
Tonga43.516.77.25.53.1
Vanuatu43.411.84.33.01.7
Average, small island states3.74.916.04.75.12.9
Fiji63.49.32.34.01.7
Papua New Guinea44.113.84.03.52.5
Sources: Country authorities’ central bank prudential data; and IMF staff estimates.Note: All numbers, unless otherwise mentioned, are percentages calculated based on average yearly data covering the period 2000–13. Papua New Guinea’s numbers are from 2000–12. The number of commercial banks is as of end-2014.

Bank-Specific Variables

In this section, we use the broad definition of interest rate spreads; that is, the quarterly net interest income as a percentage of total interest-bearing assets (called interest margins hereafter). Relationships between bank spreads and a set of selected bank-specific variables are examined. As a start, these relationships are explored visually using scatter plots, and the mechanism and intuition through which each variable could affect spreads are discussed.3

Several empirical studies find a positive relationship between operating costs and interest margins. Many banking studies, including Park and Weber (2006), Claeys and Vander Vennet (2008), Tregenna (2009), and Mirzaei, Moore, and Liu (2013), found that low operational efficiency is reflected in high bank margins. These results provide support for the structure-conduct-performance paradigm, which posits that banks can pass high operating costs almost fully on to customers in highly concentrated markets. One way to account for this is to include an index for management quality—generally measured with the ratio of noninterest expenses to total assets—as an explanatory variable. This analysis uses three disaggregated measures of operating costs: (1) salaries and wages, which are scaled by total assets and are expected to have a positive effect on spreads; (2) depreciation and occupancy costs as a ratio of net fixed assets, which are expected to have a negative impact because they may contribute to improving banks’ productivity; and (3) other costs as a ratio of total assets, which are expected to have a positive effect.4

In principle, the higher the overhead costs in the banking sector the larger the required spreads to compensate for the additional costs. Figures 17.5 and 17.6 confirm the presence of this relationship for salaries and wages and other costs. This result is consistent with the findings of Berger, Hanweck, and Humphrey (1987), who argue for the positive correlation between the relative scale of banks and agency costs. Surprisingly, spreads exhibit the same positive relationship for capital costs (see Figure 17.7); however, some period outliers seem to drive this result, which can be mitigated in the formal empirical analyses.

Figure 17.5Average Spreads and Staff Costs

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

Figure 17.6Average Spreads and Other Costs

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

Figure 17.7Average Spreads and Physical Capital Costs

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

Another important determinant of a bank’s interest margins is the scale of operations. However, the relationship between these two variables is ambiguous. On one hand, increasing the scale of operations may lead to lower average costs, which in turn can translate into smaller spreads when a bank faces competitive pressures to pass on cost savings to customers. On the other hand, a bank’s costs are, in part, affected by the risk-taking behavior of its managers. Deficient risk management functions and poor asset quality feed into higher nonperforming loans (NPLs), and borrowers can then face higher spreads. Using the growth rate of loans as a proxy of this variable, Figure 17.8 shows a negative correlation, which is consistent with the low share of loans in total assets and the moderate financial inclusion in the Solomon Islands.5

Figure 17.8Average Spreads and Scale of Operations

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

An additional bank-specific variable that should be considered in the analysis of bank spreads is the extent of risk aversion. A more risk-averse bank will hold more equity in its capital structure, and, to lower profit variability, its managers will tend to secure their deposit base by offering higher deposit rates. Accordingly, risk aversion is proxied by the ratio of equity to total assets. Figure 17.9 shows a positive correlation between bank averages of interest spreads and the risk aversion index.6

Figure 17.9Average Spreads and Risk Aversion

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

Empirical research clearly finds a positive relationship between bank interest spreads and interest credit risk, which influences margins positively, suggesting that banks add a default risk premium to loan rates. The default risk is captured through the ratio of NPLs to total loans. Intuitively, the ability to pass on the costs of NPLs to borrowers via increased margins allows commercial banks to maintain positive returns. Surprisingly, this relationship does not seem to be supported by a simple relationship between historical averages of spreads and credit risk (Figure 17.10).

Figure 17.10Average Spreads and Credit Risk

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

Sector-Specific Variables

The Herfindahl-Hirschman index (HH index) is viewed in the literature as a measure of concentration—the extent to which a few banks dominate market shares with respect to total assets, loans, or deposits. The HH index is a standard measure of consolidation in any industry and it is defined as the sum of the squared deposit, asset, or loan shares of all the banks in the market. By construction, the HH index has an upper value of 10,000 in the case of a monopolistic firm with a 100 percent share of the market; the index tends to zero in the case of a large number of firms with very small market shares.

In the present context, the HH index is measured as the sum of squares of banks’ market shares of the banking industry’s total assets, total deposits, or total loans. Generally, banks in highly concentrated markets earn monopoly rents, because they tend to collude. Collusion may result in higher rates being charged on loans and lower interest rates being paid on deposits. As in many studies presented in banking literature (for instance Goddard and others 2011), we find a positive relationship between concentration and bank spreads (Figure 17.11).

Figure 17.11Average Spreads and Market Concentration

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

Macroeconomic Variables

Clearly, an unstable and unfavorable macroeconomic and policy environment is perceived as more risky, and banks may compensate for it by requiring wider margins. To test for this assumption three main indicators are used, which are commonly employed in the literature to assess the impact of macroeconomic conditions on the interest rate spreads (Bolt and others 2012).

The first indicator is expected inflation. In an inflationary environment, bank costs generally rise, leading to higher borrowing costs for the private sector. Furthermore, high inflation is generally associated with an unstable and unpredictable economic environment. In other words, higher inflation is expected to lead to higher inflation-adjusted spreads if it causes banks to charge a risk premium. Actual inflation is used as a proxy for expected inflation in this chapter.

The second indicator is policy interest rates. Commercial banks usually use short-term deposits to finance long-term loans. This maturity transformation is an important function of commercial banks and is an important influence on pricing decisions for loans. The effect of an increase in the regulated interest rate—on central bank “Bokolo” bills—is twofold. First, this should increase the interest requested on loans, and deposit interest rates would react following the specificities of the banking market. Second, a second-round effect can take place through the reaction of the real sector: lower growth and high risks on loans. One should expect that the first-order effect will dominate. As a proxy we use the three-month Bokolo bill interest rate.

The third factor is real GDP growth. A larger economy might be expected to allow for economies of scale and greater competition, which drives down interest rate spreads. At the same time, if real GDP growth slows, banks are confronted with increased credit risk and they charge higher interest to borrowers. However, it is possible that a larger economy allows for greater specialization and deeper financial markets, in which riskier borrowers have better access to funds. The latest scenario is only expected to occur if the economy is already at high levels of private sector access to financial products, which is not the case in the Solomon Islands.

Legal and Empirical Results

Experts generally agree that the main barriers limiting credit supply in the Solomon Islands include insufficient investor protection in addition to family land tenure—which makes property rights unclear and limits the ability to use land as collateral—as well as bottlenecks in land registration. The Index of Economic Freedom is used to reflect this feature in the empirical analysis; this is very tractable since it covers 10 elements of economic freedom, some of which are particularly relevant in this framework, such as property rights, entrepreneurship, and, most important, the evolution of land ownership.7 Moreover, the strength of legal rights, which the index also captures, can measure the degree to which collateral and bankruptcy laws protect the rights of borrowers and lenders and thus facilitate lending. Figure 17.12 shows an ambiguous relationship between spreads and the Index of Economic Freedom, which is slightly positive and clearly led by some outliers in the sample.

Figure 17.12Average Spreads and Economic Freedom

Sources: Central Bank of the Solomon Islands; and IMF staff calculations.

Empirical Findings From a Pooled Regression

Methodology

The length and size of the sample pose challenges in data handling when the analysis is conducted on an individual bank. In that regard, the analysis of panel data brings additional information, reduces the phenomenon of multicollinearity of the variables, and increases the number of degrees of freedom. The latter would enhance the power of the tests and thus the degree of trust in their results.8

In what follows we estimate four versions of the model specified as:

In this equation i denotes bank i and t denotes a quarter t. The vector of bank-specific variables, Xit, contains COST_STAFF, COST_PCAPITAL, COST_OTHER, LOANS_GROWTH, RISK_ AVERSION, and CREDIT_RISK.

SPREAD is defined as net interest income as a percentage of total assets or (interest income – interest expense)/total assets. COST_STAFF stands for the cost related to salaries and wages and measured as salary and wages/total assets. COST_PCAPITAL denotes the cost of implementing new physical capital and identified as costs of occupancy and depreciation expenses/total assets. COST_ OTHER corresponds to other costs/total assets. LOANS_GROWTH is a measure of the scale of operation identified as the quarterly growth rate of loans and advances. RISK_AVERSION captures the degree of bank risk aversion measured as equity/total assets. CREDIT_RISK is defined as net NPLs to total loans. HHindex stands for the Herfindahl-Hirschman index and is defined as the sum of the squared bank loans/total loans. Macro comprises variables measuring the macroeconomic environment; specifically, we use inflation (INFLATION), Bokolo bill interest rates (INTEREST RATE), and real GDP growth (GROWTH). Legal denotes the Index of Economic Freedom (LEGAL).

In Table 17.2, model (1) exhibits only the bank-specific variables, X,t. Model (2) encompasses bank-specific and sectoral variables. Model (3) displays the macroeconomic variables, in addition to the others. Finally, model (4) extends model (3) by considering the legal and economic environment index as an additional explanatory variable.

Table 17.2Factors Explaining Interest Spreads
(1)(2)(3)(4)
Bank-Specific Variables
Loan growth−0.008−0.014***−0.010***−0.010***
(0.006)(0.004)(0.004)(0.004)
Cost of staff0.5820.762**0.531*0.557*
(0.451)(0.334)(0.284)(0.304)
Cost of physical capital0.020*0.001−0.002−0.002
(0.011)(0.010)(0.011)(0.011)
Other costs0.730**«0.659***0.560***0.558***
(0.177)(0.152)(0.151)(0.152)
Credit risk0.0090.007−0.009−0.008
(0.014)(0.010)(0.011)(0.012)
Risk aversion0.0220.0200.021*0.022*
(0.016)(0.013)(0.012)(0.012)
Industry Characteristics
HH index1.58E-03***8.10E-04***8.31E-04***
(1.72e-4)(1.99e-4)(2.38e-4)
Macroeconomic Variables
Inflation−0.008−0.006
(0.017)(0.018)
Interest rate0.090***0.086***
(0.024)(0.030)
Growth−0.049**−0.051**
(0.023)(0.024)
Legal Aspect
Legal−0.046
(0.180)
Intercept0.226−5.198***−2.410*0.207
(0.339)(0.680)(0.744)(10.238)
R-squared0.5610.8530.9140.913
Adjusted R-squared0.5090.8320.8950.892
Durbin-Watson statistic1.2181.6961.7421.749
Source:. IMF staff calculations.Note: HHI = Herfindahl-Hirschman index.*p = .10; **p = .05; ***p = .01.

Using pooled estimated generalized least squares to take into account cross-sectional heterogeneity, the alternative specifications of the model are estimated using the seemingly unrelated regression method. Standard errors and covariances are calculated with (panel-corrected) cross-section weights to obtain a robust estimate of the cross-section residual covariance matrix.

Empirical Results

A negative relationship was found between spreads and the scale of operations, which is statistically significant through the alternative specifications of the regression. This result suggests that banks in the Solomon Islands would pass on lower rates to borrowers assuming a surge in the volume of lending. In particular, an increase of 10 percent in total loans generates about a 1 percent decline in bank spreads.

The first regression results highlight the role of bank-level factors in determining banking spreads. The results shown in column 1 of Table 17.2 suggest that the main drivers of the fluctuations of spreads are related to the quality of management, proxied by the operating costs and, more particularly, to capital and other costs. The results are statistically and economically significant. And they are largely robust to the inclusion of banking sector, macro, and legal variables. Furthermore, salary and wages become more significant in explaining interest rate spreads, as suggested in columns 1, 3, and 4. According to the results of regressions, a fall in salary and wages (other) costs of 1 percentage point would be associated with a drop in interest spreads by 0.5–0.7 (0.6–0.7) percentage point. This result is robust to the model specification and remains significant regardless of the version of estimated equation.

To support the empirical result showing the high correlation between overhead costs and banking spreads, we compare the magnitude of these costs in the Solomon Islands to other countries during 2011. Figure 17.13 contrasts the overhead costs to total assets in 166 countries classified based on their income per capita adjusted by purchasing power parity. The figure shows a clear negative relationship between banks’ overhead costs and income. In addition, the Solomon Islands is among the top quintile with the highest banking costs in the sample (also higher than the average for the world, east Asia, and Pacific low-income countries).

Figure 17.13Relationship between Banks’ Overhead Costs and Income

Source: IMF staff calculations.

The finding on the effect of risk aversion is consistent with the theoretical interpretation in the sense that it positively affects spreads. For instance, an increase in the ratio of equity to total assets of 10 percent generates an increase in interest spreads by 2 percent. This result becomes significant at the 10 percent level in models (3) and (4) of Table 17.2. On the other hand, NPLs have an insignificant impact on the interest rate spread in all versions of the estimated models.

The HH index gauges the degree of market concentration and might show the opportunity to earn excess rates of return via collusion. Table 17.2 unambiguously shows that changes in the concentration of the respective banking sector are positively correlated with bank spreads. The elasticity is statistically significant in the different specifications. This result is somewhat expected given the small number of banks in the Solomon Islands.

In a highly inflationary environment this variable is predicted to be a significant determinant of profitability.9 Inflation in the Solomon Islands remained relatively low and stable during the period covered in this study, partly owing to the stability of exchange rates. Hence, the different versions of the regression do not show any evidence in favor of a close relationship between spreads and inflation.

The relatively high policy interest rates significantly explain bank spreads. According to the results of regressions, a fall in the central bank interest rates of 1 percentage point would be associated with a decline of spreads by roughly 8 basis points. It is fairly normal for a country using sterilization through issuing central bank bills—Bokolo bills in the Solomon Islands—to have high interest rates. Greater exchange rate flexibility could reduce the cost of sterilization, keep interest rates low, and hence reduce the spreads in the medium term—and Bokolo bill rates were already low at the end of the sample period.

In line with the existing literature, a strong negative elasticity is expected between overall economic activity and spreads in the banking sector. During booms, business opportunities are more available and the perceived risk is lower for commercial banks; therefore, spreads are expected to contract if growth increases. Growth is negatively affecting spreads with a statistically significant estimated elasticity of –0.06 in our sample.

The quality of the regulatory regime and economic environment, proxied by the Index of Economic Freedom, was found to have a positive effect on lowering spreads in the regression; however, the coefficient is not statistically significant; see model (4) in Table 17.2.

The impact of noninterest income on spreads is identified by introducing an additional bank-specific variable in the regression—the ratio of total noninterest income to total assets, which is highly related to foreign exchange operations in the Solomon Islands. This variable is expected to illustrate the implementation of the new Central Bank of Solomon Islands regulation in 2011 aiming to narrow the foreign exchange margin, which should force banks to engage more in the lending market and lower spreads. Although not reported in Table 17.2, the sensitivity of interest spreads to noninterest income is found to be weak and statistically not significant.10

A Further Look at Market Structure

In this section the structure of the banking sector is evaluated by identifying the impact of a change in the interest rate spread in one bank on the other banks. Formally, we estimate a vector autoregres-sion model of the form:

where

and A(L) is a lag matrix.

A Cholesky identification scheme is used assuming that the biggest bank (in terms of total assets)—Bank 1—is the one that affects the two other banks contemporaneously. Then Bank 2 is ranked second; so its decisions would affect Bank 3’s spreads immediately and those of Bank 1 with a one-quarter lag. Finally, Bank 3 would influence the two other banks only during the next quarter.

The impulse response functions of bank-specific spreads to idiosyncratic shocks, as well as the confidence intervals, are reported in Figure 17.14. The results show that the different banks change spreads in a coordinated fashion. In particular, following the first scenario (an unexpected increase of spreads by Bank 1), the two other banks react immediately through a mild but significant increase in interest spreads. It is worth noting that, following the same shock, Bank 3 decides to change interest spreads by virtually the same amount as in Bank 1. The same thing happens following a shock originating from Bank 2. On the other hand, the biggest bank does not seem to be significantly responsive to shocks in other banks’ spreads.

Figure 17.14Impulse Response Functions to Shocks in Bank-Specific Spreads

Source: IMF staff calculations.

Note: We construct bootstrap confidence intervals for impulse response functions from the structural vector autoregression. Solid blue lines correspond to 50th percentiles, and the dashed lines identify 90 percent confidence intervals.

To test the robustness of these findings, the same exercise is repeated using shocks on lending rates. Figure 17.15 shows the same pattern as in Figure 17.14, with highly correlated bank reactions to lending rate adjustments by their pairs.

Figure 17.15Impulse Response Functions to Shocks on Bank-Specific Lending Rates

Source: IMF staff calculations.

As the identification strategy cannot be empirically tested, a final exercise is conducted where the arrangement of variables is alternated and the robustness of the presence of collusion is assessed. The results are not reported here, but the impulse response functions are very similar, especially following the period where an idiosyncratic shock on spreads materializes. Consequently, we see this result as additional evidence of the significant influence of high banking market concentration on spreads, as reported in the previous section.

conclusion

Bank spreads and overhead costs are significantly and positively correlated. The scale of operations is another impediment to lowering the cost of borrowing from banks in the Solomon Islands. Besides, the results reported in this chapter also suggest that high market power and bank concentration tend to increase the possibility of collusion.

Some scope exists for increasing competition in the banking sector in the Solomon Islands, but financial deepening also requires the development of nonbank institutions. These include finance companies, foreign exchange dealers, and microcredit institutions, which have the potential to be competitive, apart from the main commercial banks. Financial inclusion initiatives, such as lowering the cost of remittances and mobile phone banking, could also help foster private sector development.

Another policy issue is the central bank lending rate. The empirical findings suggest that increases in the central bank lending rate are likely to increase net interest margins. Furthermore, a less supportive legal and economic environment in the Solomon Islands contributes to larger intermediation costs, although the result is not statistically significant. This argument is very often used to explain the limited access to bank loans. Although the empirical result is insignificant, this does not mean that policymakers should ignore potential reforms in this area.

References

    Berger, Allen N.1995. “The Profit-Structure Relationship in Banking: Tests of Market-Power and Efficient-Structure Hypotheses.Journal of Money, Credit and Banking27 (2): 40431.

    Berger, Allen N., GeraldHanweck, and DavidHumphrey.1987. “Competitive Viability in Banking: Scale, Scope, and Product Mix Economies.Journal of Monetary Economics20 (3): 50120.

    Bolt, Wilko, Leo deHaan, MarcoHoeberichts, Maarten vanOordt, and JobSwank.2012. “Bank Profitability during Recessions.Journal of Banking and Finance36 (9): 255264.

    Brock, Philip, and LilianaRojas-Suarez.2000. “Understanding the Behavior of Bank Spreads in Latin America.Journal of Development Economics63 (1): 11334.

    Claeys, Sophie, and Rudi VanderVennet.2008. “Determinants of Bank Interest Margins in Central and Eastern Europe: A Comparison with the West.Economic Systems32 (2): 197216.

    Crowley, Joe.2007. “Interest Rate Spreads in English-Speaking African Countries.IMF Working Paper 07/101, International Monetary Fund, Washington.

    Gelos, R. Gaston.2006. “Banking Spreads in Latin America.IMF Working Paper 06/44, International Monetary Fund, Washington.

    Goddard, John, HongLiu, PhilipMolyneux, and JohnWilson.2011. “The Persistence of Bank Profit.Journal ofBanking and Finance35 (11): 288190.

    Honohan, Patrick.2003. “The Accidental Tax: Inflation and the Financial Sector.In Taxation of Financial Intermediation: Theory and Practice for Emerging Economies, edited by PatrickHonohan.Washington: World Bank; New York: Oxford University Press.

    Mirzaei, Ali, TomoeMoore, and GuyLiu.2013. “Does Market Structure Matter on Banks’ Profitability and Stability? Emerging vs. Advanced Economies.Journal of Banking and Finance37 (8): 292037.

    Park, Kang H., and William L.Weber.2006. “Profitability of Korean Banks: Test of Market Structure versus Efficient Structure.Journal of Economics and Business58 (3): 22239.

    Samuel, Wendell, and LauraValderrama.2006. “The Monetary Policy Regime and Banking Spreads in Barbados.IMF Working Paper 06/211, International Monetary Fund, Washington.

    Tregenna, Fiona.2009. “The Fat Years: The Structure and Profitability of the U.S. Banking Sector in the Pre-Crisis Period.Cambridge Journal of Economics33 (4): 60932.

    Uhde, André, and UlrichHeimeshoff.2009. “Consolidation in Banking and Financial Stability in Europe: Empirical Evidence.Journal of Banking and Finance33 (7): 12991311.

This work would not have been possible without the generous cooperation and hospitality of the authorities in the Solomon Islands. The author also thanks Ezequiel Cabezon for excellent research assistance; Sami Ben Naceur, Luis Breuer, Ray Brooks, Reda Cherif, Fuad Hasanov, Hoe Ee Khor, Roger Kronenberg, Vicki Plater, Patrizia Tumbarello, Moez Souissi, and the staff of the Central Bank of Solomon Islands for their helpful comments.

See Brock and Rojas-Suarez (2000) for a detailed discussion of the definitions of interest rate margins.

The spread is widely considered an indicator of the efficiency of financial intermediation. High spreads can alter financial intermediation, because they discourage potential savers owing to low returns and increase financing costs for borrowers, reducing investment and growth opportunities (Mirzaei, Moore, and Liu 2013).

All scatter plots reflect the relationship between average spreads and the averages of the selected explanatory variables. The observations in Figures 17.5 to 17.12 represent combinations of average observations for each quarter; and therefore could be interpreted as time series.

The main reason behind disaggregating overhead costs is that their components may have a heterogeneous impact on the interest charged to borrowers (for example, staff costs versus capital costs).

The effect of a growing bank loan portfolio (and therefore size) on margins has been proved to be positive by Goddard and others (2011). They argue that banks with larger growth of loan portfolios benefit from economies of scale and, to some extent, benefit from increased market powers generating abnormally large margins. We believe that this should not happen in an economy with low financial inclusion—a very low level of bank loans to the private sector—as shown in Figure 17.8. Focusing on the time aspect justifies using growth rates of loans. In fact, all the variables in the regression, as described in the following section, are stationary.

Mirzaei, Moore, and Liu (2013) propose another explanation in which the ratio of equity to total assets is employed as a measure of capital strength. Capitalization is seen as the main source to cover loan losses. Well-capitalized banks increase their creditworthiness, which reduces their costs of funding, lowers the risk of bankruptcy, and increases their margins.

The index is published annually by the Heritage Foundation in partnership with the Wall Street Journal and measured based on 10 quantitative and qualitative factors, grouped into four broad categories, or pillars, of economic freedom: (1) rule of law (property rights, freedom from corruption); (2) limited government (fiscal freedom, government spending); (3) regulatory efficiency (business freedom, labor freedom, monetary freedom); and (4) open markets (trade freedom, investment freedom, financial freedom). Each of the 10 economic freedoms within these categories is graded on a scale of 0 to 100. A country’s overall score is derived by averaging these 10 economic freedoms, with equal weight given to each.

We run alternative bank-specific time series estimations, and the regressions seem to suffer from several problems, mainly owing to the small sample of the individual bank data.

Various studies find a positive correlation between spreads and inflation (see Honohan 2003 and Gelos 2006).

Alternatively, the coefficient of a dummy variable that captures the new policy implementation in 2011 was tested, and the result remains the same.

    Other Resources Citing This Publication