Determinants of Bank Profitability—Do Institutions, Globalization, and Global Uncertainty Matter for Banks in Island Economies? The Case of Fiji

: The objective of this study is to examine the influences of institutions, globalization, and world uncertainty on bank profitability in small developing economies. Consequently, we emphasize the significance of both bank-specific and other external factors influencing bank profitability. The empirical estimation is based on seven banks in Fiji—a small island economy—over the period 2000–2021. Together with bank-specific and macro factors, we account for institutions, globalization, and world uncertainty in analyzing the determinants of bank profitability. The study uses the fixed-effect estimation method. From the results, we observe that bank-specific variables, like the net interest margin, non-interest income, bank size, and capital adequacy ratio, are positively associated with bank profitability. Non-performing loans and credit risk are negatively associated with bank profitability. Macro variables, such as real GDP growth and remittances, have positive effects on bank profitability. Institutional factors, such as government effectiveness and voice and accountability, are positively associated with bank profitability. Regarding globalization, we find that it supports bank profitability. Global uncertainty and the Global Financial Crisis (2007–2008) are positively associated with profitability, whereas the global pandemic (COVID-19) is negatively associated. This study underscores the need to analyze the bank performance with factors beyond those reported in financial statements to derive a comprehensive understanding and appreciation of the complex nature of banking operations.


Introduction
Banks play a pivotal role in financing economic activities, thereby supporting the development of various market segments.A profitable banking sector can be both resilient to shocks (Athanasoglou et al. 2008) and, at times, benefit from antifragility that often arises from tail events 1 .One of the main roles of commercial banks is to support deposit mobilization and credit extension.Banks pay interest to deposit holders and charge interest to borrowers while ensuring a sufficient margin to support revenue and profit growth.To maximize their earnings and returns, banks organize their assets and liabilities while considering a range of factors that can affect their profitability.
Previous studies on countries' bank profitability (Short 1979;Bourke 1989;Molyneux and Thornton 1992;Demirgüç-Kunt and Huizinga 2000) mainly focused on bank-specific For bank managers and analysts aiming at profit optimization, risk management, and financial stability, examining factors beyond financial statements is essential for informed managerial decision-making.Henceforth, our study considers factors influencing bank profitability beyond those reported in financial statements.The rest of the paper is structured as follows: Section 2 presents the literature and hypotheses, Section 3 details the methodology, and Section 4 discusses the analysis and results.Finally, Section 5 presents the conclusion.

Literature Review 2
2.1.Bank-Specific Variables 2.1.1.Net Interest Margin (NIM) Net interest margin (NIM) is a critical financial indicator employed to assess the profitability of banks and other financial institutions.It represents the difference between the interest income earned by a bank and the interest expenses incurred, expressed as a percentage of the average interest-earning assets.A higher NIM indicates a bank's effectiveness in attracting low-cost deposits or charging higher interest rates on loans (Naceur 2003).However, studies have noted mixed results regarding the impact of the NIM on profitability.For instance, in the Indonesian context, Silaban (2017) and Hasan et al. (2020) observed a positive correlation, while Seenaiah et al. (2015) found that the NIM has a negative effect on bank profitability in India 3 .For our purpose, we hypothesize that H 1 : NIM is positively associated with bank profitability.

Non-Interest Income (NII)
Non-interest income (NII) represents the revenue generated by banks through sources other than interest, encompassing fees from services, such as account maintenance, transaction fees, investment banking activities, asset management fees, and trading gains.NII provides banks with an additional stream of revenue beyond traditional interest income, allowing them to diversify their revenue sources and mitigate the impact of interest rate fluctuations.By diversifying their revenue streams, banks can offset declines in interest income during periods of low interest rates.NII can also lead to efficiency gains through economies of scale and scope, improved resource allocation through internal capital markets, and the development of bank-specific competitive advantages (Elsas et al. 2010).However, NII can be more volatile than interest income, as it is often tied to market conditions and the performance of financial markets (Stiroh and Rumble 2006).During periods of economic downturns or market volatility, certain fee-based services may experience reduced demand, negatively impacting banks' profitability.This implies that some components of NII may be exposed to higher risks than interest income.
Furthermore, Ahamed (2017) observed that Indian banks with lower asset qualities tend to derive more benefits from NII compared to banks with higher asset qualities.Nisar et al. (2018) proposed that NII can serve as a proxy for revenue diversification.Analyzing 200 commercial banks in South Asian countries, they found that the total NII positively influences bank profitability and stability.However, mixed effects are observed when specific components of NII are examined.For instance, they noted that fees and commission incomes have a negative impact, while other NII components have a positive impact.Adesina (2021) analyzed 400 commercial banks in 34 African countries over the period 2005-2015 and found that higher NIIs are negatively associated with profitability.In a recent study by Mehzabin et al. (2023), it was noted that NII enhances bank profitability when interest rates are lower.We hypothesize that H 2 : NII is positively associated with bank profitability.

Bank Size (BSIZE)
Larger banks can reap the benefits of economies of scale.As banks expand, their fixed costs are distributed across a broader asset base, leading to a reduction in average costs (Aladwan 2015).Larger banks can negotiate more favorable terms with suppliers, access funds cost effectively, and exert greater influences on interest rates and fees because of their enhanced market power compared to smaller banks.These competitive advantages can bolster their profit objectives.In contrast, smaller banks can gain a competitive edge in terms of profitability by cultivating stronger relationships with local businesses and customers than large banks, granting them access to proprietary information that proves to be valuable in setting contract terms and making sound credit underwriting decisions (Regehr and Sengupta 2016).Almaqtari et al. (2019) examined 69 banks in India over the period 2008-2017 and found a positive relationship between the bank size (BSIZE) and bank profitability.Similar findings were reported by Jeris (2021) and Isayas (2022) for banks in Bangladesh and Ethiopia, respectively.However, in the case of Nepal, Tharu and Shrestha (2019) observed no significant impact of BSIZE on profitability, while Thi Thanh Thi Thanh Tran and Phan (2020) noted the negative effect of BSIZE on bank profitability in Vietnam.
Accordingly, we hypothesize that H 3 : BSIZE is positively associated with bank profitability.

Credit Risk (CRISK)
Credit risk refers to the potential loss that a bank may incur when a borrower fails to fulfill their debt obligations by the due date or upon loan maturity, potentially leading to bankruptcy if the debt is not managed (Poudel 2012).Although lending serves as a primary source of income for banks, it also poses significant threats to bank operations if not analyzed and managed sufficiently.Studies focusing on banks in South Asia (Islam and Nishiyama 2016) and African regions (Bolarinwa et al. 2019;Saleh and Afifa 2020;Abel et al. 2023) noted the negative impact of credit risk (CRISK) on bank profitability.Accordingly, we hypothesize that H 4 : CRISK is negatively associated with bank profitability.

Non-Performing Loans (NPLs)
The impact of non-performing loans (NPLs) on bank performance can be analyzed through the lens of asymmetric information theory, which posits that one party in a transactional relationship (the borrower) typically holds more information about the transaction than the other party (the bank) (Mishkin 1992).As a bank's lending decision hinges on the information provided by borrowers, inherent uncertainty regarding loan repayment and creditworthiness can emerge if the borrower's characteristics and actions are not accurately captured (Eichengreen et al. 1998).Consequently, the problem of adverse selection arises when high-quality borrowers are replaced by low-quality borrowers, which can, over time, degrade the overall quality of a bank's loan portfolios, leading to an increase in non-performing loans, asset erosion, and diminished profitability.Studies that have documented a negative effect of NPLs on bank profitability include Kingu et al. (2018) for Tanzania, Çollaku and Aliu (2021) for Kosovo, and Karim et al. (2022) and Begum and Atiq (2023) for Bangladesh.Noting the adverse effect of growing NPLs on bank profitability, we hypothesize that H 5 : NPLs are negatively associated with bank profitability.

Capital Adequacy Ratio (CAR)
A high value of the capital adequacy ratio (CAR) may indicate a bank's ability to meet its financial obligations (Rifansa and Pulungan 2022).According to Agbeja et al. (2015), the CAR reflects the bank's capital structure and can serve as a measure against widespread distress in the banking industry.A higher level of capital can support ongoing activities and enhance the potential for business expansion.Additionally, sufficient capital instills confidence in the banking system.Recent studies on banks in Nigeria (Ajayi et al. 2019), Sri Lanka (Chandrasegaran 2020), and Indonesia (Rifansa and Pulungan 2022) have established a positive correlation between the CAR and bank profitability.Henceforth, we test the hypothesis that H 6 : CAR is positively associated with bank profitability.

GDP Growth Rate (GDPG)
GDP growth (GDPG) exerts significant influences on the overall economic activity and business expansion through investment channels.A thriving economy fosters the emergence of new businesses and the expansion of existing ones, consequently augmenting the demand for loans and interest income, thereby bolstering bank profitability (Al-Jafari and Alchami 2014; Duraj and Moci 2015).Lutf and Omarkhil (2018) conducted an examination of the macroeconomic determinants of bank profitability in Pakistan and found that GDPG positively impacts bank profitability.Similar findings were observed for banks in Nigeria (Bolarinwa et al. 2019), Nepal (Singh et al. 2021), India (Ali et al. 2022), and Tunisia (Theiri and Hadoussa 2023).Hence, we hypothesize that H 7 : GDPG is positively associated with bank profitability.

Remittances (REMs)
Remittances from abroad can augment foreign exchange reserves and enhance the stability of the home country's banking system.Individuals receiving remittances often deposit these funds into their local bank accounts, expanding the bank's pool of funds and enabling the bank to extend loans and engage in investment activities.In one study on banks in Fiji, Chand et al. (2021) found a positive correlation between remittances and bank stability, and in another (Chand et al. 2023), they observed a negative association between remittances and non-performing loans, indicating that remittances support banking operations.Similarly, Ayhan and Toufaili (2021) examined the relationship between remittances and bank profitability for banks in Lebanon, and they found a positive relationship.Subsequently, we hypothesis that H 8 : REMs are positively associated with bank profitability.

Institutions
Institutions, also known as the "rules of the game" (North 1990), play a pivotal role in shaping a society's culture, behavior, and economic activities, including financial operations.Sound institutional frameworks, characterized by efficient and effective regulatory and supervisory mechanisms, strong property rights, and an effective legal system, are essential for advancing economic and financial development (Voghouei et al. 2011).A robust rule of law and a low level of corruption are hallmarks of high-quality institutions, promoting greater accountability and stability in the financial sector and facilitating financial liberalization (Chinn and Ito 2006).
Empirical studies have explored the multifaceted relationship between institutions and bank performance.In the case of Ghana, Yakubu (2019) found that a rise in corruption levels negatively impacts bank profitability.Conversely, Bougatef (2017) observed a positive association between corruption and bank profitability in Tunisia.Focusing on stability indicators, Hasan and Ashfaq (2021) analyzed 178 countries from 2000 to 2017 and found that corruption increases non-performing loans and reduces bank profitability.Etudaiye-Muhtar and Abdul-Baki (2020) investigated the impact of the institutional quality on the capital ratios of 79 listed commercial banks from 2000 to 2016, concluding that lower regulatory quality reduces bank capital.Asteriou et al. (2021) examined the effects of regulation, transparency, and corruption on bank stability and profitability for 326 banks in 19 Eurozone countries from 2005 to 2018.They found that regulation supports bank profitability and stability, while low transparency and corruption decrease bank profitability.Similarly, Nguyen (2022) examined the effects of the institutional quality and rule of law on bank deposits in European countries and found that both factors support bank funding and resilience.Athari and Bahreini (2023) concluded that political stability, regulatory quality, the rule of law, and the control of corruption positively impact Islamic banks' profitability.They further noted that regulatory settings, including their subindices, such as the extent of the disclosure and the ease of shareholder suits, negatively impact bank profitability.A study by Bashiru et al. (2023) on banks in 33 sub-Saharan African countries reported that an increase in institutional quality is positively associated with bank profitability.From these studies, it becomes clear that institutions play a critical role in shaping the financial landscape and influencing bank performance.Strong institutions, characterized by sound regulatory frameworks, robust property rights, and an effective legal system, are essential for fostering a healthy and stable financial sector.
The World Bank (2023) defines six different institutional indicators: "Control of Corruption" measures the extent to which public power is exercised for private gain, including both petty and grand forms of corruption.The indicator "Government Effectiveness" captures perceptions of the quality of public services, the quality of the civil service and the degree of its independence from political pressures, the quality of the policy formulation and implementation, and the credibility of the government's commitment to such policies."Rule of Law" captures perceptions of the extent to which agents have confidence in and abide by the laws of a country and, in particular, the quality of contract enforcement, property rights, the police, and the courts, as well as the likelihood of crime and violence."Regulatory Quality" captures perceptions of the ability of the government to formulate and implement sound policies and regulations that permit and promote private sector development."Political Stability" measures perceptions of the likelihood of political instability.The indicator "Voice and Accountability" captures perceptions of the extent to which a country's citizens are able to participate in electing their government, as well as freedom of expression, freedom of association, and free media.Noting the different components of institutions and insights from the above studies, we examine the following hypotheses: H 9 : "Corruption Control" (CC) is positively associated with bank profitability.
H 10 : "Government Effectiveness" (GE) is positively associated with bank profitability.
H 11 : "Rule of Law" (RL) is positively associated with bank profitability.
H 13 : "Political Stability" (PS) is positively associated with bank profitability.
H 14 : "Voice and Accountability" (VA) is positively associated with bank profitability.

Globalization Index
Globalization refers to the increasing interconnectedness and integration of economies, cultures, and societies worldwide (Kilic 2015).It recognizes the growing interdependence among countries in terms of trade, investment, technology, and information flow.The Globalization Index, calculated by Dreher (2006) and upgraded by Dreher et al. (2008), consists of three categories of subindices-the economic (EINDX), social (SINDX), and political (PINDX) globalization indices.Considering banks in Jordon, Guesmi et al. (2012) showed that financial globalization (measured by reductions in credit control and entry barriers and by regulatory effectiveness) is negatively associated with bank performance.For banks in Malaysia, Sufian (2012) found that social globalization (information flow, personal contacts, and cultural proximity) supports bank efficiency.Ghosh (2016) examined banks in 169 countries over the period 1998−2013 and noted that globalization reduces profitability but increases cost efficiency.Using data on commercial banks in South Africa over the period 1998-2012, Sufian and Kamarudin (2016) found positive associations of economic integration and trade and a negative association of social globalization with bank profitability.Moreover, for banks in sub-Saharan Africa, Yakubu and Bunyaminu (2022) used financial globalization and trade globalization as measures of economic globalization.They found negative and significant effects of trade and financial globalization on bank profitability.Using these studies as a guide, we investigate the following hypotheses: H 15 : EINDX is positively associated with bank profitability.
H 16 : PINDX is positively associated with bank profitability.
H 17 : SINDX is positively associated with bank profitability.

World Uncertainty Index (WUI)
The World Uncertainty Index (WUI), which quantifies the level of uncertainty in the global economy, is based on risk indicators and news (Ahir et al. 2022).A higher index indicates a greater level of uncertainty.Ashraf (2021) examined the impact of the WUI on the loan-pricing decisions of banks in 88 countries from 1998 to 2017.The study notes that a high value of the WUI (global uncertainty) is positively associated with lending rates.Bilgin et al. (2021) investigated the effects of the WUI on the default risks of conventional and Islamic banks.They noted that the relationship between the global uncertainty and default risk was not significant in Islamic banks; however, the relationship was significant for conventional banks that have high income diversification, are of a large size, and are publicly traded.Hence, we test the hypothesis that H 18 : WUI is negatively associated with bank profitability.

Structural Variables
The Global Financial Crisis (GFC) of 2007-2008 had adverse effects on many banks.The crisis was caused by a combination of factors, including the bursting of the housing price bubble in the US, excessive risk-taking by financial institutions, and the use of complex financial instruments, which reduced consumer confidence in the banking system (Prasad and Reddy 2009).Recognizing the adverse impacts of the GFC on the financial market, we examine the hypothesis that H 19 : GFC is negatively associated with bank profitability.
The COVID-19 (COVID) pandemic had negative effects on the overall economy, as evidenced by many businesses downsizing their operations, undergoing restructuring, or even shutting down.During events like COVID-19, individuals typically rely on their deposits (drawing down deposits) and, if they have loan obligations, face challenges in meeting their loan repayments.Consequently, some households and businesses end up consolidating or winding up their investments to support basic consumption and livelihood.Such actions tend to reduce a bank's revenue potential.Hence, we hypothesize that H 20 : COVID is negatively associated with bank profitability.

Studies on Banking Profitability in Fiji
Although a limited number of studies have delved into various aspects of the banking industry in island economies, like Fiji's, research on bank profitability remains scarce.For instance, Sharma and Gounder (2012) examined the determinants of net interest margins (NIMs) and noted that NIMs are positively associated with interest payments, operating costs, market power, and credit risk but negatively associated with management quality and liquidity risk.In another study, Sharma et al. (2015) found that non-interest income is the only factor that positively influences profitability.However, Krishna et al. (2021) argued that the bank size and exchange rate also support bank profitability, whereas factors like the management efficiency, a broad money supply, and the quality of the regulation reduces bank profitability.Noting the opposing results from these studies and the dearth of comprehensive studies on bank profitability in small economies, like Fiji's, we conduct this study, which considers a more extensive dataset and accounts for the roles of institutions, globalization, and global uncertainty.

The Data
This study utilizes data from seven financial institutions (FIs) over the period from 2000 to 2021, with a balanced panel.The sample comprises five commercial banks-Australia and New Zealand Banking Group (ANZ), Bank of the South Pacific (BSP), Bank of the Baroda (BOB), Westpac Banking Corporation (WBC), and Home Finance Corporation (HFC)-along with two non-bank financial institutions (NBFIs)-Merchant Finance Limited (MFL) and Credit Corporation Fiji Limited (CCFL).Bank-specific variables are sourced from disclosure statements published by the Reserve Bank of Fiji (Reserve Bank of Fiji (RBF) 2023), and variables concerning institutions and other macroeconomic aspects are obtained from Worldwide Governance Indicators and World Development Indicators (World Bank 2023).The three dimensions of the KOF Globalization Index (i.e., economic, political, and social indices) are extracted from https://datafinder.qog.gu.se/dataset/dr access on 30 March 2023 (cf.Dreher 2006;Dreher et al. 2008).The WUI index (global uncertainty) is sourced from https://worlduncertaintyindex.com/data/ access on 30 March 2023 (Ahir et al. 2022; World Uncertainty Index (WUI) 2023).In Table 1, we summarize the dependent and independent variables used for the estimations.All the data (excluding the dummy variables) are log-transformed for regression analysis 4 .

Model Specifications
Following a few recent studies (Theiri and Hadoussa 2023;Chand et al. 2021Chand et al. , 2023)), we specify the profitability model as follows: where ROA it represents bank profitability, as measured by the return on the assets of bank i at time t; α is a constant; β m contains the set of coefficients of the bank-specific factors; X m contains a set of bank-specific factors, m; θ n contains the coefficients of the respective nth globalization index, G n ; γ p contains the coefficients of the institutional factors, I p ; ω q contains the coefficients of the macro and structural (including uncertainty) variables, Z q (see Table A2); and ε ∼ N 0, σ 2 is the error term.Based on Equation ( 1), the specific model for the estimations can be expressed as follows: For a robust analysis, we estimate six models using the panel's fixed-effect method.In Model (I) we only include bank-specific variables (base model).In Model (II) we extend Model (I) with globalization.In Model (III), we extend Model (I) with globalization and institutional variables.Model (IV) represents the extended version of Model (III) with macroeconomic variables.Next, Model (V) is an extension of Model (IV) with the uncertainty index.Finally, in Model (VI) we extend Model (V) by incorporating structural variables, like "Global Financial Crisis (GFC)" and "COVID-19".Panel data may exhibit group effects, time effects, or both, which can be addressed using either fixed-or randomeffect models.Moreover, the fixed-effect regression model assumes differences in intercepts across groups or time periods (Almaqtari et al. 2019;Bolarinwa et al. 2019;Mehzabin et al. 2023).Hence, we determine the suitable estimation approach using the Hausman test (Ali and Puah 2019;Theiri and Hadoussa 2023).In this test, the rejection of the null hypothesis indicates support for the fixed-effect regression model.

Empirical Results
The descriptive statistics and correlation matrices are presented in Tables 2 and 3.The fixed-effect model is supported by the results of the Hausman test (see Table 4).There is an overall increasing trend in profits for all the banks, although ANZ, BSP, and WBC have a higher share of bank profits than others.The years 2020 and 2021 marked the episodes of lockdown that were put in place in Fiji because of the COVID-19 pandemic.During this period, we note that bank profitability declined for all the banks.In the year 2022, a positive trend in bank profitability is noted (Figure 1) 5 .fixed-effect model is supported by the results of the Hausman test (see Table 4).There is an overall increasing trend in profits for all the banks, although ANZ, BSP, and WBC have a higher share of bank profits than others.The years 2020 and 2021 marked the episodes of lockdown that were put in place in Fiji because of the COVID-19 pandemic.During this period, we note that bank profitability declined for all the banks.In the year 2022, a positive trend in bank profitability is noted (Figure 1) 5 .Table 5 presents the outcomes of the fixed estimations 6 .According to the results, the adjusted R-squared values range between 0.76 and 0.82, signifying that 80 percent of the relationship is explained by the respective models.These findings are discussed in relation to bank-specific, globalization, institutional, and macrostructural variables.It is observed that net interest margin (NIM) demonstrates a positive association with bank profitability (see Models I-VI).This is consistent with existing studies (Silaban 2017;Hasan et al. 2020).Hence, we do not reject H 1 .Moreover, the non-interest income (NII) is positive and significant at the 1 percent level (see Models I-VI), which implies that incomes from fees and commissions significantly improve bank profitability, duly supporting H 2 .This result is consistent with those in some other studies (Adesina 2021;Nisar et al. 2018;Sharma et al. 2015), which argued that in a low-interest environment, NII can be an important source of revenue.Bank size (BSIZE) is positively associated with bank profitability (see Models I-VI), implying that larger banks have higher profitability because of the high-quality assets deployed in generating consistently high incomes (Aladwan 2015;Almaqtari et al. 2019).Hence, we do not reject H 3 .Moreover, we note that the coefficient of the credit risk (CRISK), measured by the ratio of the loans to the total assets, is negative and significant (see Models I-VI).For a developing country, like Fiji, an increase in lending relative to the total assets constrains the banks' liquidity and, hence, negatively influences bank profitability-an outcome, which is consistent with those of previous studies (Abel et al. 2023;Bolarinwa et al. 2019;Islam and Nishiyama 2016).Subsequently, we do not reject H 4 .
Additionally, it is observed that Non-Performing Loans (NPLs) exhibit a negative association with bank profitability (see Models I-VI), confirming H 5 .Similar to the findings of Kingu et al. (2018), it can be argued that an escalation in non-performing loans constrains bank capital and amplifies losses, hence, exerting an adverse pressure on bank profitability.The coefficient of the Capital Adequacy Ratio (CAR) is positive and significant, suggesting that an increase in CAR positively impacts bank profitability (see Models I-VI).Hence, we do not reject H 6 .This result is consistent with findings from other developing countries' banks (Ajayi et al. 2019;Chandrasegaran 2020;Rifansa and Pulungan 2022).The coefficient of the economic growth (real GDP growth rate) is positive and significant (see Models IV-VI).Economic growth provides the necessary condition for banks to grow their investments and expand their revenue-generating activities, thus supporting bank profitability (Al-Jafari and Alchami 2014; Duraj and Moci 2015).Moreover, the coefficient of the REM is positive and significant (see Models IV-VI).Hence, remittances saved as deposits in banks tend to increase the source of funds for banks, which, in turn, supports banks' income-generating (lending and investment) activities.Accordingly, we do not reject H 7 and H 8 , respectively.Regarding institutions, we find that institution-specific variables, like Government Effectiveness (GE) and Voice and Accountability (VA), have significant and positive relationships with bank profitability (see Models III-VI).Hence, we do not reject H 10 and H 14 , respectively.Moreover, these results underscore the importance of the government, and voice and accountability, as critical drivers of bank profitability.Concerning other institutional factors, like Corruption Control (CC), Rule of Law (RL), Regulatory Quality (RQ), and Political Stability (PS), although we observe a negative relationship, these relationships are not statistically significant (see Models III-VI).The non-significant coefficients suggest either a weak (negative) effect, or these factors operate independently of bank profitability.
Regarding globalization, we observe that bank profitability demonstrates a positive association with economic (EINDX) and social (SINDX) globalization (see Models II-VI).Based on these positive and significant associations, we do not reject H 15 and H 17 , respectively.We note that political globalization is negative but not statistically significant.Economic globalization (EINDX) reflects the degree of openness in trade and financial activities.In the case of Fiji, we note that trade and financial openness supports bank profitability.With increasing globalization, banks operating in Fiji are in healthier positions to tap into international markets for investment activities and facilitate the trade of goods and services, which enables banks to increase transactions and generate more revenues.Therefore, the effective participation of banks in facilitating global trade supports bank profitability.This result is also consistent with the findings of Rahman et al.'s (2021) analysis on banks in BRICS countries.
Additionally, concerning the impact of social globalization (SINDX) on bank profitability in Fiji, a positive relationship is observed.Banks in Fiji can benefit from social globalization by proactively investing in building their reputation and brand image and by actively engaging in socially responsible initiatives.Social initiatives, such as community development projects, promoting environmental sustainability, or investing in education and healthcare, are the key areas where banks can create positive impacts.In the long-term, such initiatives can improve the welfare of society and garner customer loyalty and trust.As an example, we note that since 2010, the Bank of the South Pacific (BSP) has invested over FJD 10 million in Fiji through its Corporate Social Responsibility (CSR) program.The bank's expansive CSR program includes supporting and funding education, environmental clean-up, sporting events, and charities at community and national levels (BSP 2021).Other banks and non-financial institutions are also noted to actively participate in supporting community-level initiatives.For example, in the past, larger banks, like ANZ (ANZ 2023) and Westpac, have provided financial support in education, human capital development, and rural banking (scholarships, financial education, and rural banking support).Regarding the relationships between political globalization (PINDX) and bank profitability and between the political stability index (PS) and bank profitability, we find they are negative although not statistically significant (see Models II-VI).These results are contrary to our hypotheses H 13 and H 16 , respectively.
The World Uncertainty Index (WUI), which measures global uncertainty, exhibits a positive and statistically significant relationship with bank profitability (see Models V-VI).This positive association can be attributed to increased global uncertainty leading local investors and businesses to adopt risk-averse behaviors, hence favoring domestic sectors.Higher global uncertainty could also drive individuals and companies to secure deposits and loans and other investment opportunities from a relatively remote economy, like Fiji's.Consequently, banks may experience an influx of deposits, or bank transactions, as well as loans, thus potentially boosting revenues.Furthermore, heightened global uncertainty might prompt banks to exercise caution in lending practices, maintaining high-quality loan portfolios.All these reasons may explain the positive association between global uncertainty and bank profitability.Hence, we do reject H 18 .
Like the relationship between the WUI and bank profitability, we note that the GFC has positive and significant effects on bank profitability (see Model VI); thus, we reject H 19 .Moreover, given the remoteness and weak links of Fiji's financial sector to the rest of the world, and basic financial products and services offered by banks, this result is not surprising.Arguably, with the existing characteristics of Fiji's banking sector, events, like the GFC, tend to encourage investors to seek safer options for their investments, and even consolidate their overseas investments and funds to domestic banks, duly supporting domestic bank operations and profitability.
For the COVID-19 pandemic, we note a negative and significant association with bank profitability (see Model VI).This decline in the profitability ratio (net profit) is also captured in Figure 1, where we saw a drop in bank profitability between 2020 and 2021.The COVID-19 pandemic affected individuals, businesses, and the economy as a whole.In 2020, Fiji's real GDP fell by 15.7% (ADB 2021), and the economy witnessed relatively higher unemployment and financial hardships for both individuals and businesses.Moreover, banks faced with high chances of loan defaults were noted to be under increased pressure to implement repayment holidays and accumulated high non-performing loans-all of which deteriorated their asset quality and, hence, profitability (Badunenko et al. 2022).Hence, based on these results, we do not reject H 20 .

Conclusions
In this study, we focus on the small economy of Fiji from 2000 to 2021.We contribute to the banking literature on bank profitability in small island economies by investigating the effects of globalization, institutions, and global uncertainty.Through this study, we emphasize that the profitability and bank performance analysis should go beyond bankspecific factors.
From these results, we find that globalization, institutions, global uncertainty, and the pandemic (COVID-19) are critical factors for the sound operation and development of a robust and resilient financial sector.Moreover, because of the economy's small size and limited sectors, banks possess a constrained set of portfolios for investment.Hence, analyzing profitability by considering factors beyond financial statements aids in comprehending the intricate environment within which banks operate.
The results derived from this comprehensive model reveal that bank-specific factors, like the net interest margin, non-interest income, bank size, and capital adequacy ratio, support bank profitability, whereas factors like non-performing loans and credit risk have negative effects on bank profitability.Additionally, economic and social globalization, institutional variables (Government Effectiveness and Voice and Accountability), global uncertainty, and the Global Financial Crisis show positive associations with bank profitability.Other factors, such as economic growth and remittances, support bank profitability.According to these results, we emphasize the significance of high economic growth and exploring opportunities to maximize gains from remittance inflows.Therefore, remittances channeled through formal channels (such as banks and money transfer services) can bolster greater financial intermediation 7 .In the realm of social globalization, banks should continue to align their operations with best practices that promote social and green development projects, human capital development, technological advancement, and healthcare initiatives.
Moreover, Fiji, akin to other small countries in the region, grapples with distinctive challenges, including natural disasters, high government debt, inadequate infrastructure, reliance on limited growth sectors, and declining agriculture.This low development, coupled with events like COVID-19, intensifies adverse effects, stalling economic progress and financial sector development.Additionally, susceptibility to frequent natural disasters brings added challenges in commercializing the agricultural sector for large-scale production.Hence, banks need to factor in these events while assessing profitability and lending capacity.Fiji's heavy reliance on its tourism sector (Kumar and Stauvermann 2023) for economic growth not only limits the number of sectors available for banks to diversify their loan portfolios, but the reliance on tourism also reduces the long-term revenue potential, which becomes even more troublesome during events like COVID-19.
Because globalization emerges as a significant force shaping bank profitability, to support banking operations, the Government of Fiji should create an enabling environment to foster increased trade, attract foreign direct investments, and encourage innovative pathways of economic transactions (mobile and electronic transactions) and greater technology inclusion.
Furthermore, comprehending the dynamics between institutions and bank profitability offers valuable insights into creating specific institutional mechanisms for a stable and robust banking system in small economies.As evident from these findings, specific institutions play a vital role in shaping banking sector operations and profitability.Therefore, the government and the Reserve Bank of Fiji (central bank) should ensure a sound regulatory environment, supportive government policies, and a robust institutional infrastructure.A clear rule of law that supports investment activities, coupled with effective and efficient banking sector regulations and policies, is imperative.Fiji has experienced four political crises (coups d'état) in the past.To foster thriving banks and maintain a healthy performance level, the government must ensure political stability, uphold constitutional norms and the rule of law, promote voice and accountability, and ensure regulatory and policy certainty.The prevailing global uncertainty presents opportunities for banks in Fiji.Hence, banks should recognize this and prudently expand their operations to grow revenue and profit.
Overall, these results highlight the intricate nature of banking sector operations.As such, bank regulators, government entities, and financial managers should recognize the implications of these factors and incorporate them when formulating pertinent policies for sustainable bank operations and the holistic advancement of the financial sector.Considering economic and geopolitical circumstances, banks in small nations, such as Fiji, should meticulously analyze profitability and risk.The assessment of risk and profit should, therefore, adopt a comprehensive perspective to develop more secure strategies that bolster bank resilience and ensure sustainability.
In conclusion, we present the impacts of additional dimensions on bank profitability with certain considerations.It is important to note that these results may not be applicable to other countries, as this study centers on the banking sector of a small economy.Additionally, we acknowledge that the positive correlation between global uncertainty and the global financial crisis could stem from Fiji's limited financial integration with major centers and the small, less-advanced, and less-integrated nature of its banking sector.It is crucial to recognize that economic, political, and regulatory landscapes differ across countries and regions.Consequently, their individual or collective influences might vary for banks operating elsewhere globally.Henceforth, our study comes with its own set of limitations.First, our focus is exclusively on financial institutions engaged in deposit and lending operations.Thus, although our findings hold significance within this realm, they should not be generalized to other financial entities, such as insurance firms, pension funds, or the entire financial system.Second, although our exploration of the linear relationship between bank profit and its determinants yields valuable insights, it is essential to acknowledge the plausibility of non-linear dynamics or presence of threshold levels of key factors.Third, our sample size comprises seven banks over the period 2000-2021.Hence, the sample can be regarded as relatively small.However, we have included the seven banks in the sample because they have consistent data since 2001, available on the central bank's website.The cutoff point of our sample is also contingent upon non-bank factors, which were available up to 2021.Consequently, the small sample size may not capture the full extent of this relationship.Moreover, although we have one more bank (BRED Bank) operating in Fiji, the bank was established in 2012; hence, including BRED Bank in the balanced panel would significantly reduce the sample size.Therefore, future studies could consider these limitations with the possibility for extending this study further, as more data become available.Another important suggestion would be for the Reserve Bank of Fiji to update banks' and other financial institutions' data before 2001 because most of the financial institutions were established decades earlier.Nevertheless, within these limitations, our study presents a comprehensive set of factors, and an objective previously not explored, and offers deeper insight into bank profitability for a small island economy, like Fiji's.These  The concept of antifragility supersedes resilience and robustness and refers to conditions in which an entity gains through adverse events (Taleb 2012).
2 See Table A1 (in the Appendix A section) for a summary of the studies on the determinants of bank profitability.
3 However, the study does not provide a theoretical explanation for the negative association.We argue that it is possible that the net interest margin is sufficiently low or sufficiently high; hence, it exhibits a non-linear effect.If too low or too high, an NIM can indicate the mispricing of assets (loans) and liabilities (deposits). 4 Fiji has seven commercial banks.However, for our analysis, to ensure we have a sufficiently large sample size (2000-2021), we have excluded BRED Bank from the sample because it started operating from 2012.

5
For bank profitability, the graph demonstrates bank profitability from 2000 to 2022; however, our empirical estimation sample is from 2000 to 2021.This is because data on globalization indices are only available up to 2021.FJD 1 can be approximated as USD 0.45. 6 The results based on the ordinary least squares method are presented in the Appendix A section (see Table A2) for reference only. 7 Recently, the country experienced a surge in remittance inflows.In comparison with 2021, in 2022, remittances increased by 25 percent to FJD 1.04 billion (USD 0.52 billion) (Chand 2023).

Figure 1 .
Figure 1.Net profits in millions of FJD.Figure 1. Net profits in millions of FJD.

Figure 1 .
Figure 1.Net profits in millions of FJD.Figure 1. Net profits in millions of FJD.
Note: * Represents significance at the 0.01 level.Source: Authors' own estimation.

Table 5 .
Regression results based on fixed effects: Dependent variable is ROA.

Table A2 .
Regression results based on ordinary least squares estimation (OLS): Dependent variable is ROA.