Get 20M+ Full-Text Papers For Less Than $1.50/day. Start a 14-Day Trial for You or Your Team.

Learn More →

Is Economic Uncertainty a Risk Factor in Bank Loan Pricing Decisions? International Evidence

Is Economic Uncertainty a Risk Factor in Bank Loan Pricing Decisions? International Evidence risks Article Is Economic Uncertainty a Risk Factor in Bank Loan Pricing Decisions? International Evidence Badar Nadeem Ashraf School of Finance, Jiangxi University of Finance and Economics, Nanchang 330013, China; badarfcma@gmail.com Abstract: Uncertainty in economic environment leads economic agents to act cautiously. In this paper, we postulate that such uncertainty leads banks to charge higher interest rate on loans. Measuring aggregate country-level economic uncertainty with the World Uncertainty Index (WUI) and using a bank-level dataset from 88 countries over the period 1998–2017, we find that heightened economic uncertainty increases bank loan interest rates. Specifically, bank loan interest rates rise by 20.67 basis points with a one standard deviation increase in WUI. Our results are robust when we use alternative proxy of uncertainty, include additional controls in the model, and extend the sample size. We also observe that WUI index is better at measuring local economic uncertainty as compared to the Economic Policy Uncertainty (EPU) index. Overall, this study provides evidence that bank price in economic uncertainty is an important risk while setting interest rates on bank loans. Keywords: bank loan interest rates; economic uncertainty; political uncertainty; world uncertainty index; economic policy uncertainty 1. Introduction Citation: Ashraf, Badar Nadeem. Uncertainty about economic environment has an important bearing on the decision 2021. Is Economic Uncertainty a Risk making of economic agents at the micro-level. For instance, as uncertainty heightens, firms Factor in Bank Loan Pricing hold more cash Cheng et al. (2018); Phan et al. (2019) and reduce capital investment Decisions? International Evidence. Azzimonti (2018); Gulen and Ion (2016), inventory holdings Zeng et al. (2019), and merger Risks 9: 81. https://doi.org/ and acquisitions Bonaime et al. (2018). Likewise, households consume less and save more 10.3390/risks9050081 Aaberge et al. (2017); Giavazzi and McMahon (2012) and reduce the amount of risky assets such as stocks and bonds in their portfolios Park and Suh (2019). The financial Academic Editor: Eliana Angelini sector is also not an exception. For instance, equity investors charge a risk premium for Received: 30 December 2020 uncertainty Pastor and Veronesi (2013). Similarly, banks hoard more liquidity Ashraf (2020); Accepted: 8 April 2021 Berger et al. (2020) and lend less Bordo et al. (2016); Hu and Gong (2019); Nguyen et al. Published: 23 April 2021 (2020). In this study, we present the evidence regarding the impact of economic uncertainty on banks’ loan pricing. Publisher’s Note: MDPI stays neutral Theoretically, heightened economic uncertainty may increase bank loan pricing through with regard to jurisdictional claims in two channels: By enhancing information asymmetry between lenders and borrowers and published maps and institutional affil- due to the recessionary impact on economic activities. iations. In pioneer studies, Stiglitz and Weiss (1981) and Greenwald et al. (1984) show that information asymmetry between lenders and borrowers leads to credit rationing. More related to our study, the theoretical model of Greenwald and Stiglitz (1990) demonstrate that greater uncertainty exacerbates information asymmetry between borrowers and lenders, Copyright: © 2021 by the author. and tightens financing constraints. Recent empirical research shows that uncertainty has Licensee MDPI, Basel, Switzerland. dominant effect on the information environment of firms. For instance, Chen et al. (2018) This article is an open access article find that as uncertainty heightens, the total amount of idiosyncratic information about a distributed under the terms and firm that is available to the market decreases. In other words, firms have a propensity to conditions of the Creative Commons reduce the amount and quality of information provided to external investors. Likewise, Attribution (CC BY) license (https:// Yung and Root (2019) report that the quality of financial information deteriorates in periods creativecommons.org/licenses/by/ of higher policy uncertainty as firms tend to manage earnings more. With worsened 4.0/). Risks 2021, 9, 81. https://doi.org/10.3390/risks9050081 https://www.mdpi.com/journal/risks Risks 2021, 9, 81 2 of 17 information environment, it becomes difficult for lenders to assess the creditworthiness of borrowing firms. As a result, banks demand higher interest rate to lend. For the second channel, Ashraf and Shen (2019) argue policy uncertainty boosts bank loan interest rates by raising the default risk of borrowers. Since uncertainty shocks lead to decrease in investment, employment, household consumption, and, consequently, the GDP Baker et al. (2016); Bloom (2009); Bloom et al. (2018), the idiosyncratic dispersion in firms’ productivity Brand et al. (2019) and household incomes Bloom (2014); Li et al. (2018) increases. The adverse effect of uncertainty shock at micro-level is not limited to a few firms or households, rather it increases the variance of firms’ productivity at individual, industry, and aggregate levels Bloom (2009) and boosts household income volatility not only due to higher unemployment and less new hiring, but also because of changes in wages of those who remain employed Bloom (2014); Li et al. (2018). Thus, the uncertainty shock enhances the probability of bad state for both borrowing firms and households. In response, risk-averse banks increase average loan interest rates to cover potential loan losses. To examine the impact of economic uncertainty on bank loan pricing, we use bank- level data from 88 countries over the period 1998–2017. Following Ashraf and Shen (2019), we measure bank loan pricing with annual bank interest income to gross loans ratio. Political and economic uncertainty is measured with the world uncertainty index (WUI) of Ahir et al. (2018). We find significant positive association between WUI and bank loan interest rates. More specifically, one standard deviation increase in WUI increases bank loan rates by 20.67 basis points. We observe that our results are robust when we use an alternative proxy of political uncertainty and include additional controls in our model. This paper is different in various aspects from two related studies by Francis et al. (2014) and Ashraf and Shen (2019). In this regard, Francis et al. (2014) use data from the U.S. and examine the impact of firm-level exposure to political uncertainty on bank loan spreads. To gauge firm-level exposure to political uncertainty, they use the political uncertainty index from Baker et al. (2016). On the other hand, Ashraf and Shen (2019) employ bank-level data from 17 countries and examine the impact of economic policy uncertainty on bank loan prices. They measure economic policy uncertainty with news-based EPU index developed by Baker et al. (2016). Different from them, we examine the impact of economic uncertainty on bank loan interest rates using a large bank-level dataset from 88 countries. To gauge economic uncertainty, we use the WUI index recently developed by Ahir et al. (2018). The WUI index is calculated by counting the word uncertainty (or its variants) in the quarterly Economist Intelligence Unit (EIU) country reports, which cover country-specific politics, economic policy, the domestic economy, and foreign and trade payments events. The WUI index offers several advantages in measuring economic uncertainty. First, in contrast to the EPU index, which just measures economic policy uncertainty based on newspaper articles, the WUI index is more comprehensive and captures overall uncertainty related to economic, financial, and political trends in a country. Second, the WUI index better captures the local economic uncertainty as compared to the EPU index, which arguably measures domestic economic policy uncertainty however is more global in nature Ahir et al. (2018). Newspapers articles counted for the EPU index may also include those articles that discuss uncertainty related to international factors. Because of this, the EPU index is more likely to co-move internationally; international factors explain 36% variation in the EPU index while only 17% in the WUI index Ahir et al. (2018). Third, the WUI index for different countries is constructed based on country specific reports from same single source, which mitigates concerns about the ideological bias and consistency of the WUI and makes it easier to compare the index in levels across countries. Lastly, the WUI index is available for a large number of countries (i.e., 143 countries) in contrast to the EPU index, which is available just for 22 countries. This study offers at least two important contributions to the existing literature: First, this paper complements the literature that explores the factors affecting the bank loan pricing decisions Asquith et al. (2005); Ge et al. (2017); Huang et al. (2018); Qian and Strahan (2007); Valta (2012); Waisman (2013). For instance, these studies report that bank loan Risks 2021, 9, 81 3 of 17 interest rates incorporate the premium for borrowers’ credit quality Asquith et al. (2005), the level of competition in borrower ’s industry Valta (2012); Waisman (2013), the level of overinvestment by the borrower Ge et al. (2017), and the quality of corporate governance of borrowing firm Huang et al. (2018), among others. Extending this debate, we provide comprehensive evidence how economic uncertainty impacts bank loan interest rates. Second, this study also adds to the literature that argues that uncertainty leads to higher financing costs for corporate firms. In this regard, recent studies report that firms’ cost of equity capital Brogaard and Detzel (2015); Pastor and Veronesi (2013); Pham (2019) and bond spreads Bradley et al. (2016); Waisman et al. (2015) rise as uncertainty goes up. Complementing these findings, we show that economic uncertainty increases cost of bank financing for firms. The paper is organized as follows. Sections 2–5 present literature review, sample description, empirical model, empirical results, and conclusion, respectively. 2. Data Collection We started our sample construction by downloading the data of the WUI index developed by Ahir et al. (2018) from the website http://www.policyuncertainty.com on 20 April 2020. Country-level quarterly data of the WUI index are available for 143 countries. We collected data of other country-level financial and macroeconomic control variables from World Development Indicators (WDI) and Financial Development databases of World Bank. Next, we downloaded bank-level annual financial statements accounting data of deposit-taking financial institutions (i.e., commercial, cooperative and savings banks) from the Bankfocus (previous name was ‘Bankscope’) database over the period of 1998 to 2017. Bankfocus reports data of both active and inactive banks. To avoid any survival bias of prudent and well-managed banks, we kept both active and inactive banks in our sample. Finally, we linked annual bank-level data with annual country-level data. We dropped observations with missing values. We also dropped banks with less than 5 annual observa- tions over the whole sample period. Our final dataset consists of 34,752 annual observations of 3513 banks from 88 countries over the period from 1998 to 2017. We winsorize bank-level variables at the one percent level in both lower and upper tails to minimize the impact of outliers on empirical results. The detail about sample countries and the number of banks and annual observations from each country is given in Table 1. Table 1. Sample distribution. This table reports sample distribution and country-level mean values of dependent and main independent variables. WUI is world uncertainty index of Ahir et al. (2018). EPU index is news-based economic policy uncertainty index of Baker et al. (2016). Annual Interest Income to Sr. No. Country Banks WUI EPU Index Observations Gross Loans Ratio 1 Albania 13 128 7.84 0.16 2 Algeria 12 96 7.32 0.10 3 Angola 15 122 12.16 0.08 4 Argentina 66 863 19.14 0.32 5 Armenia 15 160 13.38 0.06 6 Australia 27 228 6.31 0.16 4.65 7 Azerbaijan 27 254 14.48 0.12 8 Bangladesh 37 250 10.86 0.07 9 Belarus 24 178 17.70 0.09 10 Bolivia 14 155 10.01 0.24 11 Botswana 10 79 14.10 0.27 12 Brazil 121 1335 23.55 0.27 4.89 13 Bulgaria 20 213 8.91 0.22 14 Burkina Faso 8 80 9.09 0.25 15 Canada 54 390 4.52 0.15 5.08 Risks 2021, 9, 81 4 of 17 Table 1. Cont. Annual Interest Income to Sr. No. Country Banks WUI EPU Index Observations Gross Loans Ratio 16 Chile 22 198 9.43 0.13 4.64 17 China 143 1234 6.10 0.09 5.10 18 Colombia 24 219 14.05 0.24 4.64 19 Congo, Dem. Rep. 6 46 12.66 0.37 20 Costa Rica 54 579 14.97 0.16 21 Cote d’Ivoire 11 89 8.12 0.27 22 Croatia 32 366 9.17 0.11 4.20 23 Czech Republic 23 232 7.63 0.17 24 Dominican Republic 52 496 20.78 0.10 25 Egypt, Arab Rep. 22 210 9.42 0.18 26 Gambia, The 3 15 18.60 0.09 27 Greece 14 131 5.59 0.12 4.66 28 Guatemala 11 80 16.76 0.18 29 Haiti 7 62 13.10 0.30 30 Honduras 5 31 15.77 0.19 31 Hong Kong SAR, C 25 250 5.94 0.10 4.85 32 Hungary 10 120 6.69 0.22 33 India 77 977 10.60 0.10 4.49 34 Indonesia 81 652 12.01 0.13 35 Iraq 2 14 12.78 0.12 36 Israel 14 212 5.45 0.21 37 Jamaica 5 43 12.57 0.14 38 Japan 640 10,111 2.51 0.18 4.63 39 Jordan 17 228 8.21 0.08 40 Kenya 31 274 13.73 0.36 41 Korea, Rep. 14 101 5.57 0.22 4.93 42 Kuwait 9 100 8.00 0.13 43 Latvia 24 258 8.21 0.16 44 Lebanon 26 210 7.15 0.28 45 Lithuania 11 120 5.22 0.12 46 Madagascar 5 42 13.28 0.16 47 Malaysia 34 211 4.39 0.13 48 Mali 9 81 8.60 0.16 49 Mexico 48 463 11.47 0.24 4.06 50 Moldova 12 113 13.14 0.29 51 Mongolia 3 18 12.81 0.16 52 Morocco 11 134 6.56 0.08 53 Namibia 9 99 11.08 0.19 54 Netherlands 19 124 6.47 0.17 4.65 55 New Zealand 16 138 6.95 0.16 56 Nicaragua 6 55 15.25 0.24 57 Niger 5 42 9.09 0.19 58 Nigeria 18 123 13.83 0.45 59 Oman 7 95 6.47 0.19 60 Pakistan 27 227 9.89 0.09 61 Panama 81 539 8.98 0.17 62 Papua New Guinea 1 5 10.38 0.08 63 Paraguay 13 80 13.26 0.22 64 Peru 19 203 13.95 0.24 65 Philippines 32 318 8.72 0.16 66 Poland 62 476 7.05 0.25 67 Qatar 10 115 5.89 0.05 68 Romania 26 276 9.97 0.17 69 Russia 505 3889 15.20 0.24 4.97 70 Senegal 13 104 7.97 0.14 71 Sierra Leone 3 21 16.87 0.12 72 Singapore 7 78 3.80 0.06 4.72 73 Slovenia 17 213 5.76 0.15 74 South Africa 22 240 12.59 0.59 Risks 2021, 9, 81 5 of 17 Table 1. Cont. Annual Interest Income to Sr. No. Country Banks WUI EPU Index Observations Gross Loans Ratio 75 Sri Lanka 15 90 11.86 0.13 76 Sweden 67 344 5.61 0.16 4.37 77 Switzerland 245 1244 3.23 0.11 78 Tanzania 25 230 13.85 0.19 79 Thailand 28 344 5.54 0.22 80 Togo 6 54 9.31 0.21 81 Uganda 17 159 17.19 0.18 82 Ukraine 70 424 14.70 0.27 83 United Arab Emirates 24 189 6.08 0.17 84 Uruguay 13 111 11.37 0.17 85 Venezuela, RB 37 469 22.22 0.25 86 Vietnam 31 242 11.58 0.13 87 Yemen, Rep. 5 35 10.21 0.08 88 Zambia 12 106 15.84 0.50 Total/mean 3513 34,752 9.06 0.19 4.73 3. Empirical Methodology For empirical analysis, we adopt the following pooled OLS model developed by Ashraf and Shen (2019). k l T1 k l Y = a + b WUI + b X + b X + e D + # . (1) i,j,t i 1 j,t å k å l å t t i,j,t i,j,t j,t k=1 l=1 t=1 where i, j, and t subscripts represent bank, country, and year, respectively. a is a constant term. # is an idiosyncratic error term. i,j,t Y, the dependent variable, represents the bank loan interest rate. Following Ashraf and Shen (2019), the bank loan interest rate is measured with annual interest income to gross loans ratio. This ratio measures the average interest rate, which banks charge on their loan portfolio in a year. a is a constant term. WUI is the main explanatory variable and stands for annual country-level economic uncertainty. X is a set of bank-level annual i,j,t control variables including return-on-equity ratio, interest-expense-to-total-liabilities ratio, operating-profit-to-total-assets ratio, non-interest-expenses-to-total-assets ratio, loan-loss- provisions-to-gross-loans ratio, loans-to-deposits ratio, and bank size. X is a set of j,t country-level variables including banking industry concentration, monetary policy rate, lending interest rate, GDP growth, inflation, developing countries dummy, and banking crisis dummy. D is a set of year dummy variables. # is an idiosyncratic error term. i,j,t WUI is the world uncertainty index developed by Ahir et al. (2018) and represents the overall uncertainty related to economic environment of a country. Ahir et al. (2018) construct the WUI index by searching the words “uncertain”, “uncertainty”, and “uncer- tainties” in EIG reports for each country and quarter. The raw count of uncertainty-related words is scaled by the total number of words in each report to make the index comparable across countries. The WUI index is available at a quarterly frequency. Since our bank-level data are annual, we averaged quarterly values of WUI to get the annual value. While setting loan interest rate, a bank considers its funding costs, expenses to provide financial service, the premium for borrowers’ risk, profit margin, the level of competition in the banking industry, its’ own position in the market, the strategies to expand in credit market, and macroeconomic factors Ashraf and Shen (2019). We add several variables to control for these confounding effects. To control for bank funding costs, we use return on equity ratio and interest expense to total liabilities ratio. Return on equity ratio measures the realized return for bank equity holders and controls for the required rate of return of bank shareholders. On the other Risks 2021, 9, 81 6 of 17 hand, interest expense to total liabilities ratio measures the interest expense paid to bank depositors and short- and long-term debt holders, and thus controls for the bank debt funding costs. The non-interest-expenses-to-total-assets ratio is included to control for bank costs to provide financial services. Likewise, return on assets (i.e., pre-impairment- operating-profit-to-total-assets ratio) is added to control for bank profit margins. To control for borrowers’ risk, we add the annual-loan-loss-provisions-to-gross-loans ratio. Loan loss provisions show the banks’ assessment of potential risks in their loan portfolios. Banks with risky loan portfolios need to book higher provisions to cover potential future loan losses. Thus, this variable controls for the average risk of all borrowers of a bank. Based on the simple cost plus loan price model, we expect that the higher the bank funding costs, costs to provide services, profit margin, and borrowers’ risk, the higher the interest rate that banks would charge on loans. We include the country-level monetary policy rate and the lending interest rate to control for cross-country differences in bank funding costs and borrowers’ risk, respectively. Monetary policy rate is the monetary policy interest rate or bank rate that the central bank of a country regularly sets to manage money supply. A tight monetary policy would increase bank funding costs. Several studies have suggested that banks respond to monetary policy changes and adjust their loan rates accordingly Becker et al. (2012); Blot and Labondance (2013); Espinosa-Vega and Rebucci (2004); Gregor and Melecký (2018). Country-level lending interest rate is the average interest rate that lenders charge on short- and medium-term loans to private sector. This rate depends on borrowers’ creditworthiness and objectives of financing. Lending interest rate would be high in countries that have higher risk. Lending interest rate may contain an average premium for cross-country differences in economic uncertainty. However, adding it as a control variable would confirm whether the WUI index captures the marginal impact of economic uncertainty on banks’ loan interest rates decisions. To control for banking industry competition, a bank’s position in market, and bank strategy towards market, we include banking industry concentration, bank size, and loans- to-deposits-ratio, respectively. The impact of banking industry concentration and bank size on loan pricing is uncertain. On the one hand, large banks in a concentrated industry enjoy economies of scale and might pass on a low cost to customers by charging lower interest rates on loans. On the other hand, they might ask higher interest rate due to substantial market power. Banks with an aggressive strategy in financial intermediation are likely to charge lower interest rates to gain market share. Recent literature suggests that uncertainty is counter-cyclical and is systematically higher in developing countries Ahir et al. (2018); Bloom (2014). Therefore, we add annual GDP Growth rate and inflation variables in Equation (1) to control for domestic business cycles. This will mitigate the concern that WUI represents the domestic business cycles. Likewise, we add the developing countries dummy variable, which equals 1 if a sample country is developing and 0 otherwise to control for countries’ income level. Since our sample period is fairly long, we add the banking crisis dummy variable in the model to control for the effect of banking crises on loan interest rates. Loan interest rates are likely to decline in crisis periods. We add time dummies to control for the effects of global business cycles. Table 2 summarizes the definitions of main variables used in this study. Risks 2021, 9, 81 7 of 17 Table 2. Variable definitions. Variable Definition Data Source Dependent variables Interest income to gross loans Equals annual interest income to gross loans ratio of each bank. Bankscope database ratio Main independent variable World uncertainty index (WUI) developed by Ahir et al. (2018). This index is constructed based on the count of uncertainty related words in the quarterly Economist Intelligence Unit (EIU) country reports. WUI is WUI Ahir et al. (2018) available at quarterly frequency. Since our bank-level data are annual, we averaged quarterly values of WUI to get annual value. Higher values of the index represent higher policy uncertainty in a country. Independent control variables (1) Bank-level This variable for each bank is measured as the annual net income to total Return on equity ratio Bankfocus database owners’ equity ratio. Interest expense to total This variable for each bank is measured as the annual interest expense on liabilities ratio interest bearing bank liabilities to total interest bearing liabilities ratio. Loan loss provisions to gross This variable for each bank is measured as the annual loan loss loans ratio provisions to bank gross loans ratio. Non-interest expenses to total This variable for each bank is measured as the annual non-interest assets ratio expenses to bank total assets ratio Operating profit to total assets This variable for each bank is measured as the annual pre-impairment ratio operating profit to bank total assets ratio. This variable for each bank is measured as the annual gross loans to total Loans to deposits ratio customers’ deposits ratio. Bank size is measured as the natural logarithm of annual total assets of Bank size each bank. (2) Industry-level Global financial Bank concentration measures the annual market share of three largest development Bank industry concentration banks in terms of total assets (i.e., equals the sum of assets of three largest database, World banks/the sum of assets of all commercial banks). Bank (3) Country-level International Monetary policy rate is the monetary policy interest rate or bank rate financial statistics, Monetary policy rate which central bank of a country regular sets to manage money supply. International Monetary Fund World Development Lending rate is defined as the country-level annual average interest rate Lending interest rate Indicators database, which banks charge on short- and medium-term loans to private sector. World Bank GDP growth rate Equals year-on-year annual GDP growth rate of each country. Inflation Equals annual percentage change in consumer prices in a country. Dummy variable equals 1 if World Bank ranks a country as developing Developing countries dummy and 0 otherwise. Dummy variable equals 1 if a country experiences a banking crisis in a Laeven and Valencia Crises dummy year and 0 otherwise. (2018) Additional variables News-based economic policy uncertainty index developed by Baker et al. (2016). This index is constructed based on the count of economic policy uncertainty-related articles published in major newspapers of each EPU index Baker et al. (2016) country. EPU index is available at monthly frequency and we average monthly values to get one annual value. We take natural log of annual values. Z-score= log (ROA + CAR)/(ROA)), where ROA is pre-impairment operating profit to total assets ratio, CAR is equity to total assets ratio, Z-score and (ROA) is standard deviation of annual values of pre-impairment Authors’ calculation operating profit to total assets ratio calculated over three-year overlapping window (i.e., 1998–2001, 1999–2002 and so on). Risks 2021, 9, 81 8 of 17 Table 2. Cont. Variable Definition Data Source A measure of, not just whether there are free and fair elections, but how responsive government is to its people. The less responsive it is, the more International Democratic Accountability likely it will fall. Even democratically elected governments can delude Country Risk Guide themselves into thinking they know what is best for the people, (ICRG) dataset regardless of clear indications to the contrary from the people. A measure of corruption within the political system that is a threat to foreign investment by distorting the economic and financial environment, Corruption reducing the efficiency of government and business by enabling people to assume positions of power through patronage rather than ability, and introducing inherent instability into the political process. Two measures comprising one risk component. Each sub-component equals half of the total. The “law” sub-component assesses the strength Law & order and impartiality of the legal system, and the “order” sub-component assesses popular observance of the law. 4. Empirical Results 4.1. Summary Statistics The summary statistics of the main variables is reported in Table 3. The interest- income-to-gross-loans ratio has a mean value of 9.06, which suggests sample banks on average have charged 9.06% interest on their loan portfolios. This value is comparable with the 7.53% interest rate reported by Ashraf and Shen (2019). The slight difference might be due to the different sample composition in this study; the number of countries in Ashraf and Shen (2019) is 17 while we have 88 countries. Similarly, their sample period is 1998– 2012 while our sample period spans from 1998 to 2017. The 7.80 standard deviation shows that the interest-income-to-gross-loans ratio possess considerable variation. The WUI, which is the main independent variable, has a mean value of 0.19 with a standard deviation of 0.13 around the mean value. Control variables also possess significant variation. Table 3. Summary statistics of main variables. Variables Observations Mean S.D. Min Max Interest income to gross loans ratio 34,752 9.06 7.80 1.14 43.42 WUI 34,752 0.19 0.13 0.00 1.34 Return on equity ratio 34,752 6.58 16.19 76.06 48.58 Interest expense to total liabilities ratio 34,752 3.52 3.84 0.00 20.04 Loan loss provisions to gross loans ratio 34,752 1.41 2.48 2.85 15.01 Non-interest expenses to total assets ratio 34,752 4.26 5.00 0.55 32.82 Operating profit to total assets ratio 34,752 1.88 2.23 3.68 11.44 Loans to deposits ratio 34,752 99.49 87.11 14.16 617.03 Bank size 34,752 14.21 2.22 9.27 20.05 Deposits interest rate 34,752 4.42 4.89 0.27 39.25 Lending interest rate 34,752 9.65 10.88 0.99 86.36 Bank industry concentration 34,752 49.73 16.90 20.85 100.00 GDP growth rate 34,752 2.79 3.91 20.60 34.47 Inflation 34,752 4.83 14.61 10.07 493.00 Developing countries dummy 34,752 0.60 0.49 0.00 1.00 Crises dummy 34,752 0.10 0.30 0.00 1.00 EPU index 20,149 4.73 0.40 3.30 5.90 Pair-wise Pearson correlation coefficients between variables are reported in Table 4. The correlation coefficients between other variables are mostly lower than 0.8 suggesting that the chances of multicollinearity in multivariate analysis are lower. Risks 2021, 9, 81 9 of 17 Table 4. Matrix of pair-wise correlations between variables. This table reports pair-wise Pearson correlations between variables. All correlations are significant at the 5% level except those in bold. Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (1) Interest income to gross loans ratio 1.00 (2) WUI 0.17 1.00 (3) Return on equity ratio 0.21 0.02 1.00 (4) Interest expense to total liabilities ratio 0.75 0.12 0.11 1.00 (5) Loan loss provisions to gross loans ratio 0.42 0.14 0.24 0.33 1.00 (6) Non-interest expenses to total assets ratio 0.64 0.15 0.04 0.45 0.44 1.00 (7) Operating profit to total assets ratio 0.57 0.11 0.50 0.39 0.38 0.34 1.00 (8) Loans to deposits ratio 0.25 0.03 0.01 0.30 0.12 0.32 0.23 1.00 (9) Bank size 0.40 0.09 0.09 0.27 0.16 0.44 0.14 0.20 1.00 (10) Deposits interest rate 0.72 0.16 0.17 0.74 0.33 0.48 0.43 0.20 0.27 1.00 (11) Lending interest rate 0.70 0.16 0.14 0.75 0.32 0.42 0.39 0.24 0.23 0.83 1.00 (12) Bank industry concentration 0.01 0.08 0.08 0.02 0.01 0.05 0.02 0.02 0.01 0.01 0.05 1.00 (13) GDP growth rate 0.11 0.12 0.23 0.09 0.08 0.10 0.16 0.04 0.05 0.00 0.02 0.05 1.00 (14) Inflation 0.29 0.06 0.13 0.21 0.12 0.21 0.20 0.06 0.16 0.36 0.24 0.00 0.13 1.00 (15) Developing countries dummy 0.62 0.14 0.22 0.59 0.27 0.44 0.44 0.24 0.25 0.62 0.54 0.03 0.32 0.25 1.00 (16) Crises dummy 0.06 0.04 0.17 0.04 0.04 0.09 0.09 0.08 0.07 0.00 0.02 0.10 0.30 0.00 0.17 1.00 Risks 2021, 9, 81 10 of 17 4.2. Policy Uncertainty and Bank Loan Interest Rates Model 1 in Table 5 reports results of baseline model, while the WUI index is added in Model 2. Table 5. Economic and political uncertainty and bank loan pricing: Main specification. This table reports results regarding the impact of economic uncertainty on bank loan interest rates. Bank loan interest rate is measured with annual ‘interest income to gross loans ratio’ and is dependent variable in all models. WUI is world uncertainty index of Ahir et al. (2018), which represents economic uncertainty. Others are control variables. Detailed definitions of variables are given in Table 2. The results are estimated with pooled OLS estimator using heteroskedasticity robust standard errors. P-values are given in parenthesis. *** represents statistical significance at 1% level. Variables Interest Income to Gross Loans Ratio Model (1) Model (2) WUI 1.591 *** (0.000) Return on equity ratio 0.025 *** 0.025 *** (0.000) (0.000) Interest expense to total liabilities ratio 0.634 *** 0.636 *** (0.000) (0.000) Loan loss provisions to gross loans ratio 0.130 *** 0.128 *** (0.000) (0.000) Non-interest expenses to total assets ratio 0.444 *** 0.443 *** (0.000) (0.000) Operating profit to total assets ratio 0.653 *** 0.645 *** (0.000) (0.000) Loans to deposits ratio 0.007 *** 0.007 *** (0.000) (0.000) Bank size 0.407 *** 0.403 *** (0.000) (0.000) Deposit interest rate 0.120 *** 0.122 *** (0.000) (0.000) Lending interest rate 0.110 *** 0.107 *** (0.000) (0.000) Bank industry concentration 0.011 *** 0.010 *** (0.000) (0.000) GDP growth rate 0.113 *** 0.124 *** (0.000) (0.000) Inflation 0.023 *** 0.023 *** (0.000) (0.000) Developing countries dummy 0.522 *** 0.479 *** (0.000) (0.000) Crises dummy 1.270 *** 1.210 *** (0.000) (0.000) Year FE Yes Yes Constant 8.508 *** 8.103 *** (0.000) (0.000) Observations 34,752 34,752 R-squared 0.781 0.782 The results of baseline model are consistent with expectation and validate our model. For instance, bank-level variables representing bank funding costs (i.e., the return-on- equity ratio and the interest-expense-to-total-liabilities ratio), bank profit margin (i.e., the operating-profit-to-total-assets ratio), bank costs to provide financial services (i.e., the non-interest-expenses-to-total-assets ratio), and the average borrowers’ risk (i.e., the loan- loss-provisions-to-total-assets ratio) all result as positive and significant suggesting that bankers set higher loan interest rates if their funding and operational costs, profit margin, and borrowers’ average risk are higher. Risks 2021, 9, 81 11 of 17 Likewise, positive results of country-level monetary policy rate and lending interest rate variables suggest banks charge higher loan interest rates in countries with tightened monetary policy and higher average risk, respectively. The negative association of banking industry concentration and bank size implies large banks in a concentrated banking indus- try benefit from economies of scale and charge lower rates on loans, results consistent with Berger et al. (2005) and Grechyna (2018). Banks with aggressive financial intermediation strategy charge lower rates as shown by results of the loans-to-deposits ratio. Similarly, banks increase loan rates in response to speculative demand due to accelerated GDP growth, while reducing loan interest rates during crises due to adverse demand shock. Finally, banks charge higher interest rates in developing countries due to poor institutional environment and higher economic and financial risks. The WUI index results as positive and significant at the 1% level in Model 2. This result indicates that banks price risks related to economic uncertainty into loan interest rates and increase bank loan interest rates in response to higher uncertainty. Economically, one standard deviation increase in the WUI index (0.13) increases bank loan rates by 0.2067 (1.59  0.13) where the mean value of bank loan rates is 9.06%. Alternatively, one standard deviation increase in WUI increases bank loan rates by 20.67 basis points. These results are comparable with Francis et al. (2014) who report a 11.9-basis-points increase and Ashraf and Shen (2019) who found a 21.84-basis-points increase in loan interest rates in response to a one-standard-deviation change in uncertainty. Together, these results confirm that economic uncertainty leads to higher loan interest rates. 4.3. Robustness Tests We perform several robustness tests to further confirm the above results. First, we use EPU index as an alternative proxy of uncertainty. Since EPU index is only available for around 20 countries, sample size drops substantially for this robustness test. As shown in Table 6, Model (1), EPU index enters positive and significant, which confirms that policy uncertainty results in higher loan interest rates. Second, since uncertainty can also boost loan interest rates raising default risk of banks Francis et al. (2014), following Ashraf and Shen (2019), we add z-score to control for bank-level idiosyncratic default risk. Z-score is calculated as follows: Z-score = log (ROA + CAR)/(ROA)) (2) where ROA is the pre-impairment-operating-protit-to-total-assets ratio, CAR is the annual- equity-to-total-assets ratio, and (ROA) is the standard deviation of annual values of the pre-impairment-operating-profit-to-total-assets ratio calculated over a three-year overlap- ping window (i.e., 1998–2001, 1999–2002, and so on). Z-score measures the distance from potential bank default, where higher values show the higher probability of bank default and vice versa. Z-score has been widely used by recent literature to measure bank default risk Ashraf (2017); Houston et al. (2010); Kanagaretnam et al. (2014). Additionally, (ROA) represents the volatility in overall bank operating income and we use it as an alternative proxy of bank income risk. Due to the three-year window, sample size decreases in regres- sions with Z-score. The WUI index still results as positive and significant after controlling for bank risk as shown in Table 6, Model (2). Consistent with expectation, risky banks charge higher rates on loans. Third, to isolate the effect of economic uncertainty from the design and quality of political institutions, which also impact financing costs Belkhir et al. (2017), we add variables from ICRG dataset to control for democratic accountability, corruption, and law and order situation. The results of WUI index largely remain similar as shown in Table 6, Models (3) to (5). At the same time, results of additional controls suggest that banks charge lower loan interest rates in countries with higher democratic accountability, lower corruption, and better law and order situation. Risks 2021, 9, 81 12 of 17 Table 6. Economic and political uncertainty and bank loan pricing: Robustness checks with additional controls. This table reports results of robustness checks regarding the impact of policy uncertainty on bank loan interest rates. Bank loan interest rate is measured with annual ‘interest income to gross loans ratio’ and is the dependent variable in all models. WUI is world uncertainty index of Ahir et al. (2018), which represents economic uncertainty. EPU index is the economic policy uncertainty index of Baker et al. (2016). Others are control variables. Detailed definitions of variables are given in Table 2. The results are estimated with pooled OLS estimator using heteroskedasticity robust standard errors. P-values are given in parenthesis. *** and ** represent statistical significance at 1% and 5% levels, respectively. Variables Interest Income to Gross Loans Ratio Model (1) Model (2) Model (3) Model (4) Model (5) WUI 1.437 *** 1.679 *** 1.528 *** 1.385 *** (0.000) (0.000) (0.000) (0.000) EPU index 0.215 ** (0.031) Return on equity ratio 0.015 *** 0.029 *** 0.026 *** 0.026 *** 0.025 *** (0.000) (0.000) (0.000) (0.000) (0.000) Interest expense to total liabilities ratio 0.514 *** 0.635 *** 0.642 *** 0.639 *** 0.637 *** (0.000) (0.000) (0.000) (0.000) (0.000) Loan loss provisions to gross loans ratio 0.190 *** 0.091 *** 0.128 *** 0.125 *** 0.123 *** (0.000) (0.002) (0.000) (0.000) (0.000) Non-interest expenses to total assets ratio 0.341 *** 0.465 *** 0.440 *** 0.442 *** 0.440 *** (0.000) (0.000) (0.000) (0.000) (0.000) Operating profit to total assets ratio 0.647 *** 0.613 *** 0.641 *** 0.642 *** 0.643 *** (0.000) (0.000) (0.000) (0.000) (0.000) Loans to deposits ratio 0.006 *** 0.007 *** 0.007 *** 0.007 *** 0.006 *** (0.000) (0.000) (0.000) (0.000) (0.000) Bank size 0.381 *** 0.349 *** 0.401 *** 0.401 *** 0.403 *** (0.000) (0.000) (0.000) (0.000) (0.000) Deposit interest rate 0.126 *** 0.159 *** 0.114 *** 0.122 *** 0.123 *** (0.001) (0.000) (0.000) (0.000) (0.000) Lending interest rate 0.129 *** 0.096 *** 0.110 *** 0.108 *** 0.096 *** (0.000) (0.000) (0.000) (0.000) (0.000) Bank industry concentration 0.016 *** 0.009 *** 0.009 *** 0.008 *** 0.009 *** (0.000) (0.000) (0.000) (0.000) (0.000) GDP growth rate 0.088 *** 0.099 *** 0.119 *** 0.127 *** 0.123 *** (0.000) (0.000) (0.000) (0.000) (0.000) Inflation 0.066 *** 0.020 *** 0.023 *** 0.023 *** 0.022 *** (0.003) (0.000) (0.000) (0.000) (0.000) Developing countries dummy 1.323 *** 0.498 *** 0.392 *** 0.150 0.100 (0.000) (0.000) (0.000) (0.147) (0.225) Crises dummy 0.983 *** 2.047 *** 1.180 *** 1.278 *** 1.147 *** (0.000) (0.000) (0.000) (0.000) (0.000) Zscore 0.142 *** (0.000) Democratic Accountability 0.121 *** (0.000) Corruption 0.179 *** (0.000) Law & order 0.320 *** (0.000) Year FE Yes Yes Yes Yes Yes Constant 7.018 *** 7.894 *** 8.628 *** 8.745 *** 10.011 *** (0.000) (0.000) (0.000) (0.000) (0.000) Observations 20,149 26,758 34,752 34,752 34,752 R-squared 0.819 0.799 0.782 0.782 0.782 Risks 2021, 9, 81 13 of 17 Fourth, one caveat with the above analysis is that the main sample, as reported in Table 1, excludes financial sectors of the US and large European countries raising concern about the international context of our study. Large European countries, as well as many other small countries, get excluded because the two country-level control variables, in- cluding Lending interest rate and Deposit interest rate, are missing for these countries in the WDI database of World Bank. To eliminate the concern whether including these countries would change the results, we re-estimate Equation (1) by excluding these two control variables. Exclusion of these two control variables from the model increases the effective sample size to 130 countries with 92,169 annual observations from the previous 88 countries and 34,752 annual observations. The extended sample also includes the large European countries, including France (2786 observations), Germany (29,864), Italy (4361), and the UK (688), among many others. As shown in Table 7, the WUI index results as positive and significant with the extended sample as well, ruling out the concern of biased results due to the sample selection. We still do not include the US in the extended sample due to the relatively large number of banks in the US, which may raise the concern of over-representation of a single country in the regression analysis. This can be considered as a shortcoming of our sample. Table 7. Economic and political uncertainty and bank loan pricing: Robustness test with extended sample size. This table reports results of robustness checks regarding the impact of policy uncertainty on bank loan interest rates. Bank loan interest rate is measured with annual ‘interest income to gross loans ratio’ and is the dependent variable in all models. WUI is the world uncertainty index of Ahir et al. (2018), which represents economic uncertainty. EPU index is the economic policy uncer- tainty index of Baker et al. (2016). Others are control variables. Two control variables, Deposit interest rate and Lending interest rate, are omitted from this regression to maximize sample size. Detailed definitions of variables are given in Table 2. The results are estimated with pooled OLS estimator using heteroskedasticity robust standard errors. P-values are given in parenthesis. *** represents statistical significance at 1% level. Variables Interest Income to Gross Loans Ratio Model (1) WUI 1.931 *** (0.000) Return on equity ratio 0.021 *** (0.000) Interest expense to total liabilities ratio 0.887 *** (0.000) Loan loss provisions to gross loans ratio 0.183 *** (0.000) Non-interest expenses to total assets ratio 0.364 *** (0.000) Operating profit to total assets ratio 0.649 *** (0.000) Loans to deposits ratio 0.008 *** (0.000) Bank size 0.322 *** (0.000) Deposit interest rate Omitted Lending interest rate Omitted Bank industry concentration 0.003 *** (0.000) GDP growth rate 0.079 *** (0.000) Inflation 0.036 *** (0.000) Risks 2021, 9, 81 14 of 17 Table 7. Cont. Variables Interest Income to Gross Loans Ratio Model (1) Developing countries dummy 1.513 *** (0.000) Crises dummy 0.226 *** (0.000) Year FE Yes Constant 7.570 *** (0.000) Observations 92,169 R-squared 0.695 4.4. Comparison of the Impact of WUI and EPU As described above, Ahir et al. (2018) argue that the WUI index better captures the local economic uncertainty as compared to the EPU index, which arguably measures domestic economic policy uncertainty however is more global in nature. To examine this, we re-estimate Equation (1) with and without time fixed-effects with both WUI and EPU indexes. Time-fixed effects in our model effectively control for global trends. For comparison, we keep only those countries for which the data of both indexes is available. As shown in Table 8, results of WUI index do not change with or without time fixed-effects. On the contrary, EPU loses significance in the model without time fixed effects. These results suggest that the impact of WUI is time independent while that of the EPU index is more likely to confound with time-fixed effects. Overall, these results are consistent with the argument of Ahir et al. (2018) that the EPU index is more global in nature. Table 8. Economic and political uncertainty and bank loan pricing: Comparison of WUI and EPU indexes. This table reports results regarding the comparison of WUI and EPU indexes. Bank loan interest rate is measured with annual ‘interest income to gross loans ratio’ and is dependent variable in all models. WUI is world uncertainty index of Ahir et al. (2018) which represents economic uncertainty. EPU index is the economic policy uncertainty index of Baker et al. (2016). Others are control variables. Detailed definitions of variables are given in Table 2. The results are estimated with pooled OLS estimator using heteroskedasticity robust standard errors. P-values are given in parenthesis. ***, **, * represent statistical significance at 1%, 5%, and 10% levels, respectively. Variables Interest Income to Gross Loans Ratio Model (1) Model (2) Model (3) Model (4) WUI 2.429 *** 2.041 *** (0.000) (0.000) EPU index 0.215 ** 0.106 (0.031) (0.135) Return on equity ratio 0.015 *** 0.015 *** 0.015 *** 0.014 *** (0.000) (0.000) (0.000) (0.000) Interest expense to total liabilities ratio 0.513 *** 0.514 *** 0.529 *** 0.533 *** (0.000) (0.000) (0.000) (0.000) Loan loss provisions to gross loans ratio 0.186 *** 0.190 *** 0.181 *** 0.187 *** (0.000) (0.000) (0.000) (0.000) Non-interest expenses to total assets ratio 0.343 *** 0.341 *** 0.337 *** 0.331 *** (0.000) (0.000) (0.000) (0.000) Operating profit to total assets ratio 0.642 *** 0.647 *** 0.654 *** 0.657 *** (0.000) (0.000) (0.000) (0.000) Loans to deposits ratio 0.006 *** 0.006 *** 0.006 *** 0.006 *** (0.000) (0.000) (0.000) (0.000) Bank size 0.369 *** 0.381 *** 0.382 *** 0.396 *** (0.000) (0.000) (0.000) (0.000) Deposit interest rate 0.147 *** 0.126 *** 0.119 *** 0.099 *** (0.000) (0.001) (0.002) (0.009) Lending interest rate 0.123 *** 0.129 *** 0.131 *** 0.138 *** (0.000) (0.000) (0.000) (0.000) Risks 2021, 9, 81 15 of 17 Table 8. Cont. Variables Interest Income to Gross Loans Ratio Model (1) Model (2) Model (3) Model (4) Bank industry concentration 0.016 *** 0.016 *** 0.019 *** 0.019 *** (0.000) (0.000) (0.000) (0.000) GDP growth rate 0.117 *** 0.088 *** 0.120 *** 0.108 *** (0.000) (0.000) (0.000) (0.000) Inflation 0.061 *** 0.066 *** 0.056 *** 0.061 *** (0.006) (0.003) (0.005) (0.002) Developing countries dummy 1.232 *** 1.323 *** 1.142 *** 1.171 *** (0.000) (0.000) (0.000) (0.000) Crises dummy 0.787 *** 0.983 *** 0.312 *** 0.379 *** (0.003) (0.000) (0.000) (0.000) Year FE Yes Yes No No Constant 7.385 *** 7.018 *** 7.798 *** 7.906 *** (0.000) (0.000) (0.000) (0.000) Observations 20,149 20,149 20,149 20,149 R-squared 0.820 0.819 0.819 0.818 There are two important points to consider while interpreting these results. First, the WUI and EPU measure different aspects of uncertainty. The EPU measures economic policy uncertainty while the WUI index represents overall economic uncertainty and is broader in definition. Usually, governments and central banks consider global trends to set their fiscal and monetary policies. On the other hand, country-specific economic events that create uncertainty are more local in nature. Second, such time dependence of the EPU index is less likely to affect the results of those empirical studies that control their models with time fixed-effects. 5. Conclusions In this paper, we examine the impact of economic uncertainty on bank loan pricing. We represent aggregate country-level economic uncertainty with the world uncertainty index (WUI) of Ahir et al. (2018). Using a cross-country bank-level dataset from 88 countries over the period 1998–2017, we find that banks increase interest rates on loans in response to heightened economic uncertainty. Specifically, a one-standard-deviation increase in WUI leads to a 20.67-basis-points increase in interest-income-to-gross-loans ratio. Our results are robust when we use the alternative proxy of uncertainty, include additional controls in our model, and use alternative estimation methods. We further find that the WUI index is better in measuring local economic uncertainty as compared to the EPU index. Our findings imply that policy uncertainty is an important risk factor in bank loan pricing and increases funding cost for firms and households. Policy uncertainty might increase in response to a crisis (i.e., financial, banking, or the pandemic as was the case with the outbreak of Covid-19 in early 2020) when a government has to choose among various possible potential policies. Though we control for the effect of crises in our regressions, future research may consider how policy uncertainty and crises interact to affect bank loan interest rates. Funding: No funding was obtained for this study. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Conflicts of Interest: The author declare no conflicts of interest. Risks 2021, 9, 81 16 of 17 Data Sharing: The data that support the findings of this study was collected from the Bankfocus database. Bankfocus is a proprietary database and doesn’t allow sharing data publicly. However, dataset used for this study can be requested from the corresponding author for research purposes. References Aaberge, Rolf, Kai Liu, and Yu Zhu. 2017. Political uncertainty and household savings. Journal of Comparative Economics 45: 154–70. [CrossRef] Ahir, Hites, Nicholas Bloom, and Davide Furceri. 2018. The World Uncertainty Index. Available online: https://ssrn.com/abstract=32 75033 (accessed on 10 August 2019). Ashraf, Badar Nadeem. 2017. Political Institutions and Bank Risk-Taking Behavior. Journal of Financial Stability 29: 13–35. [CrossRef] Ashraf, Badar Nadeem. 2020. Policy Uncertainty and Bank Liquidity Hoarding: International Evidence. Available online: http: //ssrn.com/abstract=3574193 (accessed on 19 April 2021). Ashraf, Badar Nadeem, and Yinjie Shen. 2019. Economic policy uncertainty and banks’ loan pricing. Journal of Financial Stability 44: 100695. [CrossRef] Asquith, Paul, Anne Beatty, and Joseph Weber. 2005. Performance pricing in bank debt contracts. Journal of Accounting and Economics 40: 101–28. [CrossRef] Azzimonti, Marina. 2018. Partisan conflict and private investment. Journal of Monetary Economics 93: 114–31. [CrossRef] Baker, Scott R., Nicholas Bloom, and Steven J. Davis. 2016. Measuring economic policy uncertainty. Quarterly Journal of Economics 131: 1593–636. [CrossRef] Becker, Ralf, Denise R. Osborn, and Dilem Yildirim. 2012. A threshold cointegration analysis of interest rate pass-through to UK mortgage rates. Economic Modelling 29: 2504–13. [CrossRef] Belkhir, Mohamed, Narjess Boubakri, and Jocelyn Grira. 2017. Political risk and the cost of capital in the MENA region. Emerging Markets Review 33: 155–72. [CrossRef] Berger, Allen N., Nathan H. Miller, Mitchell A. Petersen, Raghuram G. Rajan, and Jeremy C. Stein. 2005. Does function follow organizational form? Evidence from the lending practices of large and small banks. Journal of Financial Economics 76: 237–69. [CrossRef] Berger, Allen N., Omrane Guedhami, Hugh Hoikwang Kim, and Xinming Li. 2020. Economic policy uncertainty and bank liquidity hoarding. Journal of Financial Intermediation, 100893. [CrossRef] Bloom, Nicholas. 2009. The impact of uncertainty shocks. Econometrica 77: 623–85. Bloom, Nicholas. 2014. Fluctuations in Uncertainty. Journal of Economic Perspectives 28: 153–76. [CrossRef] Bloom, Nicholas, Max Floetotto, Nir Jaimovich, Itay Saporta-Eksten, and Stephen J. Terry. 2018. Really Uncertain Business Cycles. Econometrica 86: 1031–65. [CrossRef] Blot, Christophe, and Fabien Labondance. 2013. Business lending rate pass-through in the Enrozone: Monetary policy transmission before and after the financial crash. Economics Bulletin 2: 973–85. Bonaime, Alice, Huseyin Gulen, and Mihai Ion. 2018. Does policy uncertainty affect mergers and acquisitions? Journal of Financial Economics 129: 531–58. [CrossRef] Bordo, Michael D., John V. Duca, and Christoffer Koch. 2016. Economic policy uncertainty and the credit channel: Aggregate and bank level U.S. evidence over several decades. Journal of Financial Stability 26: 90–106. [CrossRef] Bradley, Daniel, Christos Pantzalis, and Xiaojing Yuan. 2016. Policy risk, corporate political strategies, and the cost of debt. Journal of Corporate Finance 40: 254–75. [CrossRef] Brand, Thomas, Marlène Isoré, and Fabien Tripier. 2019. Uncertainty shocks and firm creation: Search and monitoring in the credit market. Journal of Economic Dynamics and Control 99: 19–53. [CrossRef] Brogaard, Jonathan, and Andrew Detzel. 2015. The Asset-Pricing Implications of Government Economic Policy Uncertainty. Manage- ment Science 61: 3–18. [CrossRef] Chen, Yunsen, Deqiu Chen, Weimin Wang, and Dengjin Zheng. 2018. Political uncertainty and firms’ information environment: Evidence from China. Journal of Accounting and Public Policy 37: 39–64. [CrossRef] Cheng, Chak Hung Jack, Ching-Wai Jeremy Chiu, William B. Hankins, and Anna-Leigh Stone. 2018. Partisan Conflict, Policy Uncertainty and Aggregate Corporate Cash Holdings. Journal of Macroeconomics 58: 78–90. [CrossRef] Espinosa-Vega, Marco A., and Alessandro Rebucci. 2004. Retail bank interest rate pass-through: Is Chile atypical? Central banking, analysis, and economic policies book series. Banking Market Structure and Monetary Policy 7: 147–82. Francis, Bill B., Iftekhar Hasan, and Yun Zhu. 2014. Political uncertainty and bank loan contracting. Journal of Empirical Finance 29: 281–86. [CrossRef] Ge, Wenxia, Tony Kang, Gerald J. Lobo, and Byron Y. Song. 2017. Investment decisions and bank loan contracting. Asian Review of Accounting 25: 262–87. [CrossRef] Giavazzi, Francesco, and Michael McMahon. 2012. Policy uncertainty and household savings. Review of Economics and Statistics 94: 517–31. [CrossRef] Grechyna, D. 2018. Firm size, bank size, and financial development. Journal of Economic Dynamics and Control 97: 19–37. [CrossRef] Greenwald, Bruce C., and Joseph E. Stiglitz. 1990. Macroeconomic models with equity and credit rationing. In Asymmetric Information, Corporate Finance, and Investment. Chicago: University of Chicago Press, pp. 15–42. Risks 2021, 9, 81 17 of 17 Greenwald, Bruce C., Joseph E. Stiglitz, and Andrew Weiss. 1984. Informational imperfections in the capital market and macro- economic fluctuations. National Bureau of Economic Research. Available online: https://www.nber.org/system/files/working_ papers/w1335/w1335.pdf (accessed on 10 August 2019). Gregor, Jir ˇí, and Martin Melecký. 2018. The pass-through of monetary policy rate to lending rates: The role of macro-financial factors. Economic Modelling 73: 71–88. [CrossRef] Gulen, Huseyin, and Mihai Ion. 2016. Policy uncertainty and corporate investment. Review of Financial Studies 29: 523–64. [CrossRef] Houston, Joel F., Chen Lin, Ping Lin, and Yue Ma. 2010. Creditor rights, information sharing, and bank risk taking. Journal of Financial Economics 96: 485–512. [CrossRef] Hu, Shiwei, and Di Gong. 2019. Economic policy uncertainty, prudential regulation and bank lending. Finance Research Letters 29: 373–78. [CrossRef] Huang, Henry He, Gerald J. Lobo, Chong Wang, and Jian Zhou. 2018. Do Banks Price Independent Directors Attention? Journal of Financial and Quantitative Analysis 53: 1755–80. [CrossRef] Kanagaretnam, Kiridaran, Chee Yeow Lim, and Gerald J. Lobo. 2014. Influence of National Culture on Accounting Conservatism and Risk-Taking in the Banking Industry. Accounting Review 89: 1115–49. [CrossRef] Laeven, Luc, and Fabian Valencia. 2018. Systemic Banking Crises Revisited. IMF Working Paper. WP/18/206. Washington: International Monetary Fund. Li, Xiang, Bibo Liu, and Xuan Tian. 2018. Policy Uncertainty and Household Credit Access: Evidence from Peer-to-Peer Crowdfunding. Available online: https://ssrn.com/abstract=3084388 (accessed on 10 August 2019). Nguyen, Canh Phuc, Thai-Ha Le, and Thanh Dinh Su. 2020. Economic policy uncertainty and credit growth: Evidence from a global sample. Research in International Business and Finance 51: 101118. [CrossRef] Park, Jin Seok, and Donghyun Suh. 2019. Uncertainty and household portfolio choice: Evidence from South Korea. Economics Letters 180: 21–24. [CrossRef] Pastor, Lubos, and Pietro Veronesi. 2013. Political uncertainty and risk premia. Journal of Financial Economics 110: 520–45. [CrossRef] Pham, Anh Viet. 2019. Political risk and cost of equity: The mediating role of political connections. Journal of Corporate Finance 56: 64–87. [CrossRef] Phan, Hieu V., Nam H. Nguyen, and Hien T. Nguyen. 2019. Policy uncertainty and firm cash holdings. Journal of Business Research 95: 71–82. [CrossRef] Qian, Jun, and Philip E. Strahan. 2007. How laws and institutions shape financial contracts: The case of bank loans. The Journal of Finance 62: 2803–34. [CrossRef] Stiglitz, Joseph E., and Andrew Weiss. 1981. Credit Rationing in Markets with Imperfect Information. The American Economic Review 71: 393–410. Valta, Philip. 2012. Competition and the cost of debt. Journal of Financial Economics 105: 661–82. [CrossRef] Waisman, Maya. 2013. Product market competition and the cost of bank loans: Evidence from state antitakeover laws. Journal of Banking & Finance 37: 4721–37. Waisman, Maya, Pengfei Ye, and Yun Zhu. 2015. The effect of political uncertainty on the cost of corporate debt. Journal of Financial Stability 16: 106–17. [CrossRef] Yung, Kenneth, and Andrew Root. 2019. Policy uncertainty and earnings management: International evidence. Journal of Business Research 100: 255–67. [CrossRef] Zeng, Jianyu, Teng Zhong, and Fan He. 2019. Economic policy uncertainty and corporate inventory holdings: Evidence from China. Accounting and Finance 60: 1727–57. [CrossRef] http://www.deepdyve.com/assets/images/DeepDyve-Logo-lg.png Risks Multidisciplinary Digital Publishing Institute

Is Economic Uncertainty a Risk Factor in Bank Loan Pricing Decisions? International Evidence

Ashraf, Badar Nadeem
Risks , Volume 9 (5) – Apr 23, 2021
Download PDF
17 pages

Loading next page...
 
/lp/multidisciplinary-digital-publishing-institute/is-economic-uncertainty-a-risk-factor-in-bank-loan-pricing-decisions-cfkfH4gH3M

References

References for this paper are not available at this time. We will be adding them shortly, thank you for your patience.

Publisher
Multidisciplinary Digital Publishing Institute
Copyright
© 1996-2021 MDPI (Basel, Switzerland) unless otherwise stated Disclaimer The statements, opinions and data contained in the journals are solely those of the individual authors and contributors and not of the publisher and the editor(s). MDPI stays neutral with regard to jurisdictional claims in published maps and institutional affiliations. Terms and Conditions Privacy Policy
ISSN
2227-9091
DOI
10.3390/risks9050081
Publisher site
See Article on Publisher Site

Abstract

risks Article Is Economic Uncertainty a Risk Factor in Bank Loan Pricing Decisions? International Evidence Badar Nadeem Ashraf School of Finance, Jiangxi University of Finance and Economics, Nanchang 330013, China; badarfcma@gmail.com Abstract: Uncertainty in economic environment leads economic agents to act cautiously. In this paper, we postulate that such uncertainty leads banks to charge higher interest rate on loans. Measuring aggregate country-level economic uncertainty with the World Uncertainty Index (WUI) and using a bank-level dataset from 88 countries over the period 1998–2017, we find that heightened economic uncertainty increases bank loan interest rates. Specifically, bank loan interest rates rise by 20.67 basis points with a one standard deviation increase in WUI. Our results are robust when we use alternative proxy of uncertainty, include additional controls in the model, and extend the sample size. We also observe that WUI index is better at measuring local economic uncertainty as compared to the Economic Policy Uncertainty (EPU) index. Overall, this study provides evidence that bank price in economic uncertainty is an important risk while setting interest rates on bank loans. Keywords: bank loan interest rates; economic uncertainty; political uncertainty; world uncertainty index; economic policy uncertainty 1. Introduction Citation: Ashraf, Badar Nadeem. Uncertainty about economic environment has an important bearing on the decision 2021. Is Economic Uncertainty a Risk making of economic agents at the micro-level. For instance, as uncertainty heightens, firms Factor in Bank Loan Pricing hold more cash Cheng et al. (2018); Phan et al. (2019) and reduce capital investment Decisions? International Evidence. Azzimonti (2018); Gulen and Ion (2016), inventory holdings Zeng et al. (2019), and merger Risks 9: 81. https://doi.org/ and acquisitions Bonaime et al. (2018). Likewise, households consume less and save more 10.3390/risks9050081 Aaberge et al. (2017); Giavazzi and McMahon (2012) and reduce the amount of risky assets such as stocks and bonds in their portfolios Park and Suh (2019). The financial Academic Editor: Eliana Angelini sector is also not an exception. For instance, equity investors charge a risk premium for Received: 30 December 2020 uncertainty Pastor and Veronesi (2013). Similarly, banks hoard more liquidity Ashraf (2020); Accepted: 8 April 2021 Berger et al. (2020) and lend less Bordo et al. (2016); Hu and Gong (2019); Nguyen et al. Published: 23 April 2021 (2020). In this study, we present the evidence regarding the impact of economic uncertainty on banks’ loan pricing. Publisher’s Note: MDPI stays neutral Theoretically, heightened economic uncertainty may increase bank loan pricing through with regard to jurisdictional claims in two channels: By enhancing information asymmetry between lenders and borrowers and published maps and institutional affil- due to the recessionary impact on economic activities. iations. In pioneer studies, Stiglitz and Weiss (1981) and Greenwald et al. (1984) show that information asymmetry between lenders and borrowers leads to credit rationing. More related to our study, the theoretical model of Greenwald and Stiglitz (1990) demonstrate that greater uncertainty exacerbates information asymmetry between borrowers and lenders, Copyright: © 2021 by the author. and tightens financing constraints. Recent empirical research shows that uncertainty has Licensee MDPI, Basel, Switzerland. dominant effect on the information environment of firms. For instance, Chen et al. (2018) This article is an open access article find that as uncertainty heightens, the total amount of idiosyncratic information about a distributed under the terms and firm that is available to the market decreases. In other words, firms have a propensity to conditions of the Creative Commons reduce the amount and quality of information provided to external investors. Likewise, Attribution (CC BY) license (https:// Yung and Root (2019) report that the quality of financial information deteriorates in periods creativecommons.org/licenses/by/ of higher policy uncertainty as firms tend to manage earnings more. With worsened 4.0/). Risks 2021, 9, 81. https://doi.org/10.3390/risks9050081 https://www.mdpi.com/journal/risks Risks 2021, 9, 81 2 of 17 information environment, it becomes difficult for lenders to assess the creditworthiness of borrowing firms. As a result, banks demand higher interest rate to lend. For the second channel, Ashraf and Shen (2019) argue policy uncertainty boosts bank loan interest rates by raising the default risk of borrowers. Since uncertainty shocks lead to decrease in investment, employment, household consumption, and, consequently, the GDP Baker et al. (2016); Bloom (2009); Bloom et al. (2018), the idiosyncratic dispersion in firms’ productivity Brand et al. (2019) and household incomes Bloom (2014); Li et al. (2018) increases. The adverse effect of uncertainty shock at micro-level is not limited to a few firms or households, rather it increases the variance of firms’ productivity at individual, industry, and aggregate levels Bloom (2009) and boosts household income volatility not only due to higher unemployment and less new hiring, but also because of changes in wages of those who remain employed Bloom (2014); Li et al. (2018). Thus, the uncertainty shock enhances the probability of bad state for both borrowing firms and households. In response, risk-averse banks increase average loan interest rates to cover potential loan losses. To examine the impact of economic uncertainty on bank loan pricing, we use bank- level data from 88 countries over the period 1998–2017. Following Ashraf and Shen (2019), we measure bank loan pricing with annual bank interest income to gross loans ratio. Political and economic uncertainty is measured with the world uncertainty index (WUI) of Ahir et al. (2018). We find significant positive association between WUI and bank loan interest rates. More specifically, one standard deviation increase in WUI increases bank loan rates by 20.67 basis points. We observe that our results are robust when we use an alternative proxy of political uncertainty and include additional controls in our model. This paper is different in various aspects from two related studies by Francis et al. (2014) and Ashraf and Shen (2019). In this regard, Francis et al. (2014) use data from the U.S. and examine the impact of firm-level exposure to political uncertainty on bank loan spreads. To gauge firm-level exposure to political uncertainty, they use the political uncertainty index from Baker et al. (2016). On the other hand, Ashraf and Shen (2019) employ bank-level data from 17 countries and examine the impact of economic policy uncertainty on bank loan prices. They measure economic policy uncertainty with news-based EPU index developed by Baker et al. (2016). Different from them, we examine the impact of economic uncertainty on bank loan interest rates using a large bank-level dataset from 88 countries. To gauge economic uncertainty, we use the WUI index recently developed by Ahir et al. (2018). The WUI index is calculated by counting the word uncertainty (or its variants) in the quarterly Economist Intelligence Unit (EIU) country reports, which cover country-specific politics, economic policy, the domestic economy, and foreign and trade payments events. The WUI index offers several advantages in measuring economic uncertainty. First, in contrast to the EPU index, which just measures economic policy uncertainty based on newspaper articles, the WUI index is more comprehensive and captures overall uncertainty related to economic, financial, and political trends in a country. Second, the WUI index better captures the local economic uncertainty as compared to the EPU index, which arguably measures domestic economic policy uncertainty however is more global in nature Ahir et al. (2018). Newspapers articles counted for the EPU index may also include those articles that discuss uncertainty related to international factors. Because of this, the EPU index is more likely to co-move internationally; international factors explain 36% variation in the EPU index while only 17% in the WUI index Ahir et al. (2018). Third, the WUI index for different countries is constructed based on country specific reports from same single source, which mitigates concerns about the ideological bias and consistency of the WUI and makes it easier to compare the index in levels across countries. Lastly, the WUI index is available for a large number of countries (i.e., 143 countries) in contrast to the EPU index, which is available just for 22 countries. This study offers at least two important contributions to the existing literature: First, this paper complements the literature that explores the factors affecting the bank loan pricing decisions Asquith et al. (2005); Ge et al. (2017); Huang et al. (2018); Qian and Strahan (2007); Valta (2012); Waisman (2013). For instance, these studies report that bank loan Risks 2021, 9, 81 3 of 17 interest rates incorporate the premium for borrowers’ credit quality Asquith et al. (2005), the level of competition in borrower ’s industry Valta (2012); Waisman (2013), the level of overinvestment by the borrower Ge et al. (2017), and the quality of corporate governance of borrowing firm Huang et al. (2018), among others. Extending this debate, we provide comprehensive evidence how economic uncertainty impacts bank loan interest rates. Second, this study also adds to the literature that argues that uncertainty leads to higher financing costs for corporate firms. In this regard, recent studies report that firms’ cost of equity capital Brogaard and Detzel (2015); Pastor and Veronesi (2013); Pham (2019) and bond spreads Bradley et al. (2016); Waisman et al. (2015) rise as uncertainty goes up. Complementing these findings, we show that economic uncertainty increases cost of bank financing for firms. The paper is organized as follows. Sections 2–5 present literature review, sample description, empirical model, empirical results, and conclusion, respectively. 2. Data Collection We started our sample construction by downloading the data of the WUI index developed by Ahir et al. (2018) from the website http://www.policyuncertainty.com on 20 April 2020. Country-level quarterly data of the WUI index are available for 143 countries. We collected data of other country-level financial and macroeconomic control variables from World Development Indicators (WDI) and Financial Development databases of World Bank. Next, we downloaded bank-level annual financial statements accounting data of deposit-taking financial institutions (i.e., commercial, cooperative and savings banks) from the Bankfocus (previous name was ‘Bankscope’) database over the period of 1998 to 2017. Bankfocus reports data of both active and inactive banks. To avoid any survival bias of prudent and well-managed banks, we kept both active and inactive banks in our sample. Finally, we linked annual bank-level data with annual country-level data. We dropped observations with missing values. We also dropped banks with less than 5 annual observa- tions over the whole sample period. Our final dataset consists of 34,752 annual observations of 3513 banks from 88 countries over the period from 1998 to 2017. We winsorize bank-level variables at the one percent level in both lower and upper tails to minimize the impact of outliers on empirical results. The detail about sample countries and the number of banks and annual observations from each country is given in Table 1. Table 1. Sample distribution. This table reports sample distribution and country-level mean values of dependent and main independent variables. WUI is world uncertainty index of Ahir et al. (2018). EPU index is news-based economic policy uncertainty index of Baker et al. (2016). Annual Interest Income to Sr. No. Country Banks WUI EPU Index Observations Gross Loans Ratio 1 Albania 13 128 7.84 0.16 2 Algeria 12 96 7.32 0.10 3 Angola 15 122 12.16 0.08 4 Argentina 66 863 19.14 0.32 5 Armenia 15 160 13.38 0.06 6 Australia 27 228 6.31 0.16 4.65 7 Azerbaijan 27 254 14.48 0.12 8 Bangladesh 37 250 10.86 0.07 9 Belarus 24 178 17.70 0.09 10 Bolivia 14 155 10.01 0.24 11 Botswana 10 79 14.10 0.27 12 Brazil 121 1335 23.55 0.27 4.89 13 Bulgaria 20 213 8.91 0.22 14 Burkina Faso 8 80 9.09 0.25 15 Canada 54 390 4.52 0.15 5.08 Risks 2021, 9, 81 4 of 17 Table 1. Cont. Annual Interest Income to Sr. No. Country Banks WUI EPU Index Observations Gross Loans Ratio 16 Chile 22 198 9.43 0.13 4.64 17 China 143 1234 6.10 0.09 5.10 18 Colombia 24 219 14.05 0.24 4.64 19 Congo, Dem. Rep. 6 46 12.66 0.37 20 Costa Rica 54 579 14.97 0.16 21 Cote d’Ivoire 11 89 8.12 0.27 22 Croatia 32 366 9.17 0.11 4.20 23 Czech Republic 23 232 7.63 0.17 24 Dominican Republic 52 496 20.78 0.10 25 Egypt, Arab Rep. 22 210 9.42 0.18 26 Gambia, The 3 15 18.60 0.09 27 Greece 14 131 5.59 0.12 4.66 28 Guatemala 11 80 16.76 0.18 29 Haiti 7 62 13.10 0.30 30 Honduras 5 31 15.77 0.19 31 Hong Kong SAR, C 25 250 5.94 0.10 4.85 32 Hungary 10 120 6.69 0.22 33 India 77 977 10.60 0.10 4.49 34 Indonesia 81 652 12.01 0.13 35 Iraq 2 14 12.78 0.12 36 Israel 14 212 5.45 0.21 37 Jamaica 5 43 12.57 0.14 38 Japan 640 10,111 2.51 0.18 4.63 39 Jordan 17 228 8.21 0.08 40 Kenya 31 274 13.73 0.36 41 Korea, Rep. 14 101 5.57 0.22 4.93 42 Kuwait 9 100 8.00 0.13 43 Latvia 24 258 8.21 0.16 44 Lebanon 26 210 7.15 0.28 45 Lithuania 11 120 5.22 0.12 46 Madagascar 5 42 13.28 0.16 47 Malaysia 34 211 4.39 0.13 48 Mali 9 81 8.60 0.16 49 Mexico 48 463 11.47 0.24 4.06 50 Moldova 12 113 13.14 0.29 51 Mongolia 3 18 12.81 0.16 52 Morocco 11 134 6.56 0.08 53 Namibia 9 99 11.08 0.19 54 Netherlands 19 124 6.47 0.17 4.65 55 New Zealand 16 138 6.95 0.16 56 Nicaragua 6 55 15.25 0.24 57 Niger 5 42 9.09 0.19 58 Nigeria 18 123 13.83 0.45 59 Oman 7 95 6.47 0.19 60 Pakistan 27 227 9.89 0.09 61 Panama 81 539 8.98 0.17 62 Papua New Guinea 1 5 10.38 0.08 63 Paraguay 13 80 13.26 0.22 64 Peru 19 203 13.95 0.24 65 Philippines 32 318 8.72 0.16 66 Poland 62 476 7.05 0.25 67 Qatar 10 115 5.89 0.05 68 Romania 26 276 9.97 0.17 69 Russia 505 3889 15.20 0.24 4.97 70 Senegal 13 104 7.97 0.14 71 Sierra Leone 3 21 16.87 0.12 72 Singapore 7 78 3.80 0.06 4.72 73 Slovenia 17 213 5.76 0.15 74 South Africa 22 240 12.59 0.59 Risks 2021, 9, 81 5 of 17 Table 1. Cont. Annual Interest Income to Sr. No. Country Banks WUI EPU Index Observations Gross Loans Ratio 75 Sri Lanka 15 90 11.86 0.13 76 Sweden 67 344 5.61 0.16 4.37 77 Switzerland 245 1244 3.23 0.11 78 Tanzania 25 230 13.85 0.19 79 Thailand 28 344 5.54 0.22 80 Togo 6 54 9.31 0.21 81 Uganda 17 159 17.19 0.18 82 Ukraine 70 424 14.70 0.27 83 United Arab Emirates 24 189 6.08 0.17 84 Uruguay 13 111 11.37 0.17 85 Venezuela, RB 37 469 22.22 0.25 86 Vietnam 31 242 11.58 0.13 87 Yemen, Rep. 5 35 10.21 0.08 88 Zambia 12 106 15.84 0.50 Total/mean 3513 34,752 9.06 0.19 4.73 3. Empirical Methodology For empirical analysis, we adopt the following pooled OLS model developed by Ashraf and Shen (2019). k l T1 k l Y = a + b WUI + b X + b X + e D + # . (1) i,j,t i 1 j,t å k å l å t t i,j,t i,j,t j,t k=1 l=1 t=1 where i, j, and t subscripts represent bank, country, and year, respectively. a is a constant term. # is an idiosyncratic error term. i,j,t Y, the dependent variable, represents the bank loan interest rate. Following Ashraf and Shen (2019), the bank loan interest rate is measured with annual interest income to gross loans ratio. This ratio measures the average interest rate, which banks charge on their loan portfolio in a year. a is a constant term. WUI is the main explanatory variable and stands for annual country-level economic uncertainty. X is a set of bank-level annual i,j,t control variables including return-on-equity ratio, interest-expense-to-total-liabilities ratio, operating-profit-to-total-assets ratio, non-interest-expenses-to-total-assets ratio, loan-loss- provisions-to-gross-loans ratio, loans-to-deposits ratio, and bank size. X is a set of j,t country-level variables including banking industry concentration, monetary policy rate, lending interest rate, GDP growth, inflation, developing countries dummy, and banking crisis dummy. D is a set of year dummy variables. # is an idiosyncratic error term. i,j,t WUI is the world uncertainty index developed by Ahir et al. (2018) and represents the overall uncertainty related to economic environment of a country. Ahir et al. (2018) construct the WUI index by searching the words “uncertain”, “uncertainty”, and “uncer- tainties” in EIG reports for each country and quarter. The raw count of uncertainty-related words is scaled by the total number of words in each report to make the index comparable across countries. The WUI index is available at a quarterly frequency. Since our bank-level data are annual, we averaged quarterly values of WUI to get the annual value. While setting loan interest rate, a bank considers its funding costs, expenses to provide financial service, the premium for borrowers’ risk, profit margin, the level of competition in the banking industry, its’ own position in the market, the strategies to expand in credit market, and macroeconomic factors Ashraf and Shen (2019). We add several variables to control for these confounding effects. To control for bank funding costs, we use return on equity ratio and interest expense to total liabilities ratio. Return on equity ratio measures the realized return for bank equity holders and controls for the required rate of return of bank shareholders. On the other Risks 2021, 9, 81 6 of 17 hand, interest expense to total liabilities ratio measures the interest expense paid to bank depositors and short- and long-term debt holders, and thus controls for the bank debt funding costs. The non-interest-expenses-to-total-assets ratio is included to control for bank costs to provide financial services. Likewise, return on assets (i.e., pre-impairment- operating-profit-to-total-assets ratio) is added to control for bank profit margins. To control for borrowers’ risk, we add the annual-loan-loss-provisions-to-gross-loans ratio. Loan loss provisions show the banks’ assessment of potential risks in their loan portfolios. Banks with risky loan portfolios need to book higher provisions to cover potential future loan losses. Thus, this variable controls for the average risk of all borrowers of a bank. Based on the simple cost plus loan price model, we expect that the higher the bank funding costs, costs to provide services, profit margin, and borrowers’ risk, the higher the interest rate that banks would charge on loans. We include the country-level monetary policy rate and the lending interest rate to control for cross-country differences in bank funding costs and borrowers’ risk, respectively. Monetary policy rate is the monetary policy interest rate or bank rate that the central bank of a country regularly sets to manage money supply. A tight monetary policy would increase bank funding costs. Several studies have suggested that banks respond to monetary policy changes and adjust their loan rates accordingly Becker et al. (2012); Blot and Labondance (2013); Espinosa-Vega and Rebucci (2004); Gregor and Melecký (2018). Country-level lending interest rate is the average interest rate that lenders charge on short- and medium-term loans to private sector. This rate depends on borrowers’ creditworthiness and objectives of financing. Lending interest rate would be high in countries that have higher risk. Lending interest rate may contain an average premium for cross-country differences in economic uncertainty. However, adding it as a control variable would confirm whether the WUI index captures the marginal impact of economic uncertainty on banks’ loan interest rates decisions. To control for banking industry competition, a bank’s position in market, and bank strategy towards market, we include banking industry concentration, bank size, and loans- to-deposits-ratio, respectively. The impact of banking industry concentration and bank size on loan pricing is uncertain. On the one hand, large banks in a concentrated industry enjoy economies of scale and might pass on a low cost to customers by charging lower interest rates on loans. On the other hand, they might ask higher interest rate due to substantial market power. Banks with an aggressive strategy in financial intermediation are likely to charge lower interest rates to gain market share. Recent literature suggests that uncertainty is counter-cyclical and is systematically higher in developing countries Ahir et al. (2018); Bloom (2014). Therefore, we add annual GDP Growth rate and inflation variables in Equation (1) to control for domestic business cycles. This will mitigate the concern that WUI represents the domestic business cycles. Likewise, we add the developing countries dummy variable, which equals 1 if a sample country is developing and 0 otherwise to control for countries’ income level. Since our sample period is fairly long, we add the banking crisis dummy variable in the model to control for the effect of banking crises on loan interest rates. Loan interest rates are likely to decline in crisis periods. We add time dummies to control for the effects of global business cycles. Table 2 summarizes the definitions of main variables used in this study. Risks 2021, 9, 81 7 of 17 Table 2. Variable definitions. Variable Definition Data Source Dependent variables Interest income to gross loans Equals annual interest income to gross loans ratio of each bank. Bankscope database ratio Main independent variable World uncertainty index (WUI) developed by Ahir et al. (2018). This index is constructed based on the count of uncertainty related words in the quarterly Economist Intelligence Unit (EIU) country reports. WUI is WUI Ahir et al. (2018) available at quarterly frequency. Since our bank-level data are annual, we averaged quarterly values of WUI to get annual value. Higher values of the index represent higher policy uncertainty in a country. Independent control variables (1) Bank-level This variable for each bank is measured as the annual net income to total Return on equity ratio Bankfocus database owners’ equity ratio. Interest expense to total This variable for each bank is measured as the annual interest expense on liabilities ratio interest bearing bank liabilities to total interest bearing liabilities ratio. Loan loss provisions to gross This variable for each bank is measured as the annual loan loss loans ratio provisions to bank gross loans ratio. Non-interest expenses to total This variable for each bank is measured as the annual non-interest assets ratio expenses to bank total assets ratio Operating profit to total assets This variable for each bank is measured as the annual pre-impairment ratio operating profit to bank total assets ratio. This variable for each bank is measured as the annual gross loans to total Loans to deposits ratio customers’ deposits ratio. Bank size is measured as the natural logarithm of annual total assets of Bank size each bank. (2) Industry-level Global financial Bank concentration measures the annual market share of three largest development Bank industry concentration banks in terms of total assets (i.e., equals the sum of assets of three largest database, World banks/the sum of assets of all commercial banks). Bank (3) Country-level International Monetary policy rate is the monetary policy interest rate or bank rate financial statistics, Monetary policy rate which central bank of a country regular sets to manage money supply. International Monetary Fund World Development Lending rate is defined as the country-level annual average interest rate Lending interest rate Indicators database, which banks charge on short- and medium-term loans to private sector. World Bank GDP growth rate Equals year-on-year annual GDP growth rate of each country. Inflation Equals annual percentage change in consumer prices in a country. Dummy variable equals 1 if World Bank ranks a country as developing Developing countries dummy and 0 otherwise. Dummy variable equals 1 if a country experiences a banking crisis in a Laeven and Valencia Crises dummy year and 0 otherwise. (2018) Additional variables News-based economic policy uncertainty index developed by Baker et al. (2016). This index is constructed based on the count of economic policy uncertainty-related articles published in major newspapers of each EPU index Baker et al. (2016) country. EPU index is available at monthly frequency and we average monthly values to get one annual value. We take natural log of annual values. Z-score= log (ROA + CAR)/(ROA)), where ROA is pre-impairment operating profit to total assets ratio, CAR is equity to total assets ratio, Z-score and (ROA) is standard deviation of annual values of pre-impairment Authors’ calculation operating profit to total assets ratio calculated over three-year overlapping window (i.e., 1998–2001, 1999–2002 and so on). Risks 2021, 9, 81 8 of 17 Table 2. Cont. Variable Definition Data Source A measure of, not just whether there are free and fair elections, but how responsive government is to its people. The less responsive it is, the more International Democratic Accountability likely it will fall. Even democratically elected governments can delude Country Risk Guide themselves into thinking they know what is best for the people, (ICRG) dataset regardless of clear indications to the contrary from the people. A measure of corruption within the political system that is a threat to foreign investment by distorting the economic and financial environment, Corruption reducing the efficiency of government and business by enabling people to assume positions of power through patronage rather than ability, and introducing inherent instability into the political process. Two measures comprising one risk component. Each sub-component equals half of the total. The “law” sub-component assesses the strength Law & order and impartiality of the legal system, and the “order” sub-component assesses popular observance of the law. 4. Empirical Results 4.1. Summary Statistics The summary statistics of the main variables is reported in Table 3. The interest- income-to-gross-loans ratio has a mean value of 9.06, which suggests sample banks on average have charged 9.06% interest on their loan portfolios. This value is comparable with the 7.53% interest rate reported by Ashraf and Shen (2019). The slight difference might be due to the different sample composition in this study; the number of countries in Ashraf and Shen (2019) is 17 while we have 88 countries. Similarly, their sample period is 1998– 2012 while our sample period spans from 1998 to 2017. The 7.80 standard deviation shows that the interest-income-to-gross-loans ratio possess considerable variation. The WUI, which is the main independent variable, has a mean value of 0.19 with a standard deviation of 0.13 around the mean value. Control variables also possess significant variation. Table 3. Summary statistics of main variables. Variables Observations Mean S.D. Min Max Interest income to gross loans ratio 34,752 9.06 7.80 1.14 43.42 WUI 34,752 0.19 0.13 0.00 1.34 Return on equity ratio 34,752 6.58 16.19 76.06 48.58 Interest expense to total liabilities ratio 34,752 3.52 3.84 0.00 20.04 Loan loss provisions to gross loans ratio 34,752 1.41 2.48 2.85 15.01 Non-interest expenses to total assets ratio 34,752 4.26 5.00 0.55 32.82 Operating profit to total assets ratio 34,752 1.88 2.23 3.68 11.44 Loans to deposits ratio 34,752 99.49 87.11 14.16 617.03 Bank size 34,752 14.21 2.22 9.27 20.05 Deposits interest rate 34,752 4.42 4.89 0.27 39.25 Lending interest rate 34,752 9.65 10.88 0.99 86.36 Bank industry concentration 34,752 49.73 16.90 20.85 100.00 GDP growth rate 34,752 2.79 3.91 20.60 34.47 Inflation 34,752 4.83 14.61 10.07 493.00 Developing countries dummy 34,752 0.60 0.49 0.00 1.00 Crises dummy 34,752 0.10 0.30 0.00 1.00 EPU index 20,149 4.73 0.40 3.30 5.90 Pair-wise Pearson correlation coefficients between variables are reported in Table 4. The correlation coefficients between other variables are mostly lower than 0.8 suggesting that the chances of multicollinearity in multivariate analysis are lower. Risks 2021, 9, 81 9 of 17 Table 4. Matrix of pair-wise correlations between variables. This table reports pair-wise Pearson correlations between variables. All correlations are significant at the 5% level except those in bold. Variables (1) (2) (3) (4) (5) (6) (7) (8) (9) (10) (11) (12) (13) (14) (15) (16) (1) Interest income to gross loans ratio 1.00 (2) WUI 0.17 1.00 (3) Return on equity ratio 0.21 0.02 1.00 (4) Interest expense to total liabilities ratio 0.75 0.12 0.11 1.00 (5) Loan loss provisions to gross loans ratio 0.42 0.14 0.24 0.33 1.00 (6) Non-interest expenses to total assets ratio 0.64 0.15 0.04 0.45 0.44 1.00 (7) Operating profit to total assets ratio 0.57 0.11 0.50 0.39 0.38 0.34 1.00 (8) Loans to deposits ratio 0.25 0.03 0.01 0.30 0.12 0.32 0.23 1.00 (9) Bank size 0.40 0.09 0.09 0.27 0.16 0.44 0.14 0.20 1.00 (10) Deposits interest rate 0.72 0.16 0.17 0.74 0.33 0.48 0.43 0.20 0.27 1.00 (11) Lending interest rate 0.70 0.16 0.14 0.75 0.32 0.42 0.39 0.24 0.23 0.83 1.00 (12) Bank industry concentration 0.01 0.08 0.08 0.02 0.01 0.05 0.02 0.02 0.01 0.01 0.05 1.00 (13) GDP growth rate 0.11 0.12 0.23 0.09 0.08 0.10 0.16 0.04 0.05 0.00 0.02 0.05 1.00 (14) Inflation 0.29 0.06 0.13 0.21 0.12 0.21 0.20 0.06 0.16 0.36 0.24 0.00 0.13 1.00 (15) Developing countries dummy 0.62 0.14 0.22 0.59 0.27 0.44 0.44 0.24 0.25 0.62 0.54 0.03 0.32 0.25 1.00 (16) Crises dummy 0.06 0.04 0.17 0.04 0.04 0.09 0.09 0.08 0.07 0.00 0.02 0.10 0.30 0.00 0.17 1.00 Risks 2021, 9, 81 10 of 17 4.2. Policy Uncertainty and Bank Loan Interest Rates Model 1 in Table 5 reports results of baseline model, while the WUI index is added in Model 2. Table 5. Economic and political uncertainty and bank loan pricing: Main specification. This table reports results regarding the impact of economic uncertainty on bank loan interest rates. Bank loan interest rate is measured with annual ‘interest income to gross loans ratio’ and is dependent variable in all models. WUI is world uncertainty index of Ahir et al. (2018), which represents economic uncertainty. Others are control variables. Detailed definitions of variables are given in Table 2. The results are estimated with pooled OLS estimator using heteroskedasticity robust standard errors. P-values are given in parenthesis. *** represents statistical significance at 1% level. Variables Interest Income to Gross Loans Ratio Model (1) Model (2) WUI 1.591 *** (0.000) Return on equity ratio 0.025 *** 0.025 *** (0.000) (0.000) Interest expense to total liabilities ratio 0.634 *** 0.636 *** (0.000) (0.000) Loan loss provisions to gross loans ratio 0.130 *** 0.128 *** (0.000) (0.000) Non-interest expenses to total assets ratio 0.444 *** 0.443 *** (0.000) (0.000) Operating profit to total assets ratio 0.653 *** 0.645 *** (0.000) (0.000) Loans to deposits ratio 0.007 *** 0.007 *** (0.000) (0.000) Bank size 0.407 *** 0.403 *** (0.000) (0.000) Deposit interest rate 0.120 *** 0.122 *** (0.000) (0.000) Lending interest rate 0.110 *** 0.107 *** (0.000) (0.000) Bank industry concentration 0.011 *** 0.010 *** (0.000) (0.000) GDP growth rate 0.113 *** 0.124 *** (0.000) (0.000) Inflation 0.023 *** 0.023 *** (0.000) (0.000) Developing countries dummy 0.522 *** 0.479 *** (0.000) (0.000) Crises dummy 1.270 *** 1.210 *** (0.000) (0.000) Year FE Yes Yes Constant 8.508 *** 8.103 *** (0.000) (0.000) Observations 34,752 34,752 R-squared 0.781 0.782 The results of baseline model are consistent with expectation and validate our model. For instance, bank-level variables representing bank funding costs (i.e., the return-on- equity ratio and the interest-expense-to-total-liabilities ratio), bank profit margin (i.e., the operating-profit-to-total-assets ratio), bank costs to provide financial services (i.e., the non-interest-expenses-to-total-assets ratio), and the average borrowers’ risk (i.e., the loan- loss-provisions-to-total-assets ratio) all result as positive and significant suggesting that bankers set higher loan interest rates if their funding and operational costs, profit margin, and borrowers’ average risk are higher. Risks 2021, 9, 81 11 of 17 Likewise, positive results of country-level monetary policy rate and lending interest rate variables suggest banks charge higher loan interest rates in countries with tightened monetary policy and higher average risk, respectively. The negative association of banking industry concentration and bank size implies large banks in a concentrated banking indus- try benefit from economies of scale and charge lower rates on loans, results consistent with Berger et al. (2005) and Grechyna (2018). Banks with aggressive financial intermediation strategy charge lower rates as shown by results of the loans-to-deposits ratio. Similarly, banks increase loan rates in response to speculative demand due to accelerated GDP growth, while reducing loan interest rates during crises due to adverse demand shock. Finally, banks charge higher interest rates in developing countries due to poor institutional environment and higher economic and financial risks. The WUI index results as positive and significant at the 1% level in Model 2. This result indicates that banks price risks related to economic uncertainty into loan interest rates and increase bank loan interest rates in response to higher uncertainty. Economically, one standard deviation increase in the WUI index (0.13) increases bank loan rates by 0.2067 (1.59  0.13) where the mean value of bank loan rates is 9.06%. Alternatively, one standard deviation increase in WUI increases bank loan rates by 20.67 basis points. These results are comparable with Francis et al. (2014) who report a 11.9-basis-points increase and Ashraf and Shen (2019) who found a 21.84-basis-points increase in loan interest rates in response to a one-standard-deviation change in uncertainty. Together, these results confirm that economic uncertainty leads to higher loan interest rates. 4.3. Robustness Tests We perform several robustness tests to further confirm the above results. First, we use EPU index as an alternative proxy of uncertainty. Since EPU index is only available for around 20 countries, sample size drops substantially for this robustness test. As shown in Table 6, Model (1), EPU index enters positive and significant, which confirms that policy uncertainty results in higher loan interest rates. Second, since uncertainty can also boost loan interest rates raising default risk of banks Francis et al. (2014), following Ashraf and Shen (2019), we add z-score to control for bank-level idiosyncratic default risk. Z-score is calculated as follows: Z-score = log (ROA + CAR)/(ROA)) (2) where ROA is the pre-impairment-operating-protit-to-total-assets ratio, CAR is the annual- equity-to-total-assets ratio, and (ROA) is the standard deviation of annual values of the pre-impairment-operating-profit-to-total-assets ratio calculated over a three-year overlap- ping window (i.e., 1998–2001, 1999–2002, and so on). Z-score measures the distance from potential bank default, where higher values show the higher probability of bank default and vice versa. Z-score has been widely used by recent literature to measure bank default risk Ashraf (2017); Houston et al. (2010); Kanagaretnam et al. (2014). Additionally, (ROA) represents the volatility in overall bank operating income and we use it as an alternative proxy of bank income risk. Due to the three-year window, sample size decreases in regres- sions with Z-score. The WUI index still results as positive and significant after controlling for bank risk as shown in Table 6, Model (2). Consistent with expectation, risky banks charge higher rates on loans. Third, to isolate the effect of economic uncertainty from the design and quality of political institutions, which also impact financing costs Belkhir et al. (2017), we add variables from ICRG dataset to control for democratic accountability, corruption, and law and order situation. The results of WUI index largely remain similar as shown in Table 6, Models (3) to (5). At the same time, results of additional controls suggest that banks charge lower loan interest rates in countries with higher democratic accountability, lower corruption, and better law and order situation. Risks 2021, 9, 81 12 of 17 Table 6. Economic and political uncertainty and bank loan pricing: Robustness checks with additional controls. This table reports results of robustness checks regarding the impact of policy uncertainty on bank loan interest rates. Bank loan interest rate is measured with annual ‘interest income to gross loans ratio’ and is the dependent variable in all models. WUI is world uncertainty index of Ahir et al. (2018), which represents economic uncertainty. EPU index is the economic policy uncertainty index of Baker et al. (2016). Others are control variables. Detailed definitions of variables are given in Table 2. The results are estimated with pooled OLS estimator using heteroskedasticity robust standard errors. P-values are given in parenthesis. *** and ** represent statistical significance at 1% and 5% levels, respectively. Variables Interest Income to Gross Loans Ratio Model (1) Model (2) Model (3) Model (4) Model (5) WUI 1.437 *** 1.679 *** 1.528 *** 1.385 *** (0.000) (0.000) (0.000) (0.000) EPU index 0.215 ** (0.031) Return on equity ratio 0.015 *** 0.029 *** 0.026 *** 0.026 *** 0.025 *** (0.000) (0.000) (0.000) (0.000) (0.000) Interest expense to total liabilities ratio 0.514 *** 0.635 *** 0.642 *** 0.639 *** 0.637 *** (0.000) (0.000) (0.000) (0.000) (0.000) Loan loss provisions to gross loans ratio 0.190 *** 0.091 *** 0.128 *** 0.125 *** 0.123 *** (0.000) (0.002) (0.000) (0.000) (0.000) Non-interest expenses to total assets ratio 0.341 *** 0.465 *** 0.440 *** 0.442 *** 0.440 *** (0.000) (0.000) (0.000) (0.000) (0.000) Operating profit to total assets ratio 0.647 *** 0.613 *** 0.641 *** 0.642 *** 0.643 *** (0.000) (0.000) (0.000) (0.000) (0.000) Loans to deposits ratio 0.006 *** 0.007 *** 0.007 *** 0.007 *** 0.006 *** (0.000) (0.000) (0.000) (0.000) (0.000) Bank size 0.381 *** 0.349 *** 0.401 *** 0.401 *** 0.403 *** (0.000) (0.000) (0.000) (0.000) (0.000) Deposit interest rate 0.126 *** 0.159 *** 0.114 *** 0.122 *** 0.123 *** (0.001) (0.000) (0.000) (0.000) (0.000) Lending interest rate 0.129 *** 0.096 *** 0.110 *** 0.108 *** 0.096 *** (0.000) (0.000) (0.000) (0.000) (0.000) Bank industry concentration 0.016 *** 0.009 *** 0.009 *** 0.008 *** 0.009 *** (0.000) (0.000) (0.000) (0.000) (0.000) GDP growth rate 0.088 *** 0.099 *** 0.119 *** 0.127 *** 0.123 *** (0.000) (0.000) (0.000) (0.000) (0.000) Inflation 0.066 *** 0.020 *** 0.023 *** 0.023 *** 0.022 *** (0.003) (0.000) (0.000) (0.000) (0.000) Developing countries dummy 1.323 *** 0.498 *** 0.392 *** 0.150 0.100 (0.000) (0.000) (0.000) (0.147) (0.225) Crises dummy 0.983 *** 2.047 *** 1.180 *** 1.278 *** 1.147 *** (0.000) (0.000) (0.000) (0.000) (0.000) Zscore 0.142 *** (0.000) Democratic Accountability 0.121 *** (0.000) Corruption 0.179 *** (0.000) Law & order 0.320 *** (0.000) Year FE Yes Yes Yes Yes Yes Constant 7.018 *** 7.894 *** 8.628 *** 8.745 *** 10.011 *** (0.000) (0.000) (0.000) (0.000) (0.000) Observations 20,149 26,758 34,752 34,752 34,752 R-squared 0.819 0.799 0.782 0.782 0.782 Risks 2021, 9, 81 13 of 17 Fourth, one caveat with the above analysis is that the main sample, as reported in Table 1, excludes financial sectors of the US and large European countries raising concern about the international context of our study. Large European countries, as well as many other small countries, get excluded because the two country-level control variables, in- cluding Lending interest rate and Deposit interest rate, are missing for these countries in the WDI database of World Bank. To eliminate the concern whether including these countries would change the results, we re-estimate Equation (1) by excluding these two control variables. Exclusion of these two control variables from the model increases the effective sample size to 130 countries with 92,169 annual observations from the previous 88 countries and 34,752 annual observations. The extended sample also includes the large European countries, including France (2786 observations), Germany (29,864), Italy (4361), and the UK (688), among many others. As shown in Table 7, the WUI index results as positive and significant with the extended sample as well, ruling out the concern of biased results due to the sample selection. We still do not include the US in the extended sample due to the relatively large number of banks in the US, which may raise the concern of over-representation of a single country in the regression analysis. This can be considered as a shortcoming of our sample. Table 7. Economic and political uncertainty and bank loan pricing: Robustness test with extended sample size. This table reports results of robustness checks regarding the impact of policy uncertainty on bank loan interest rates. Bank loan interest rate is measured with annual ‘interest income to gross loans ratio’ and is the dependent variable in all models. WUI is the world uncertainty index of Ahir et al. (2018), which represents economic uncertainty. EPU index is the economic policy uncer- tainty index of Baker et al. (2016). Others are control variables. Two control variables, Deposit interest rate and Lending interest rate, are omitted from this regression to maximize sample size. Detailed definitions of variables are given in Table 2. The results are estimated with pooled OLS estimator using heteroskedasticity robust standard errors. P-values are given in parenthesis. *** represents statistical significance at 1% level. Variables Interest Income to Gross Loans Ratio Model (1) WUI 1.931 *** (0.000) Return on equity ratio 0.021 *** (0.000) Interest expense to total liabilities ratio 0.887 *** (0.000) Loan loss provisions to gross loans ratio 0.183 *** (0.000) Non-interest expenses to total assets ratio 0.364 *** (0.000) Operating profit to total assets ratio 0.649 *** (0.000) Loans to deposits ratio 0.008 *** (0.000) Bank size 0.322 *** (0.000) Deposit interest rate Omitted Lending interest rate Omitted Bank industry concentration 0.003 *** (0.000) GDP growth rate 0.079 *** (0.000) Inflation 0.036 *** (0.000) Risks 2021, 9, 81 14 of 17 Table 7. Cont. Variables Interest Income to Gross Loans Ratio Model (1) Developing countries dummy 1.513 *** (0.000) Crises dummy 0.226 *** (0.000) Year FE Yes Constant 7.570 *** (0.000) Observations 92,169 R-squared 0.695 4.4. Comparison of the Impact of WUI and EPU As described above, Ahir et al. (2018) argue that the WUI index better captures the local economic uncertainty as compared to the EPU index, which arguably measures domestic economic policy uncertainty however is more global in nature. To examine this, we re-estimate Equation (1) with and without time fixed-effects with both WUI and EPU indexes. Time-fixed effects in our model effectively control for global trends. For comparison, we keep only those countries for which the data of both indexes is available. As shown in Table 8, results of WUI index do not change with or without time fixed-effects. On the contrary, EPU loses significance in the model without time fixed effects. These results suggest that the impact of WUI is time independent while that of the EPU index is more likely to confound with time-fixed effects. Overall, these results are consistent with the argument of Ahir et al. (2018) that the EPU index is more global in nature. Table 8. Economic and political uncertainty and bank loan pricing: Comparison of WUI and EPU indexes. This table reports results regarding the comparison of WUI and EPU indexes. Bank loan interest rate is measured with annual ‘interest income to gross loans ratio’ and is dependent variable in all models. WUI is world uncertainty index of Ahir et al. (2018) which represents economic uncertainty. EPU index is the economic policy uncertainty index of Baker et al. (2016). Others are control variables. Detailed definitions of variables are given in Table 2. The results are estimated with pooled OLS estimator using heteroskedasticity robust standard errors. P-values are given in parenthesis. ***, **, * represent statistical significance at 1%, 5%, and 10% levels, respectively. Variables Interest Income to Gross Loans Ratio Model (1) Model (2) Model (3) Model (4) WUI 2.429 *** 2.041 *** (0.000) (0.000) EPU index 0.215 ** 0.106 (0.031) (0.135) Return on equity ratio 0.015 *** 0.015 *** 0.015 *** 0.014 *** (0.000) (0.000) (0.000) (0.000) Interest expense to total liabilities ratio 0.513 *** 0.514 *** 0.529 *** 0.533 *** (0.000) (0.000) (0.000) (0.000) Loan loss provisions to gross loans ratio 0.186 *** 0.190 *** 0.181 *** 0.187 *** (0.000) (0.000) (0.000) (0.000) Non-interest expenses to total assets ratio 0.343 *** 0.341 *** 0.337 *** 0.331 *** (0.000) (0.000) (0.000) (0.000) Operating profit to total assets ratio 0.642 *** 0.647 *** 0.654 *** 0.657 *** (0.000) (0.000) (0.000) (0.000) Loans to deposits ratio 0.006 *** 0.006 *** 0.006 *** 0.006 *** (0.000) (0.000) (0.000) (0.000) Bank size 0.369 *** 0.381 *** 0.382 *** 0.396 *** (0.000) (0.000) (0.000) (0.000) Deposit interest rate 0.147 *** 0.126 *** 0.119 *** 0.099 *** (0.000) (0.001) (0.002) (0.009) Lending interest rate 0.123 *** 0.129 *** 0.131 *** 0.138 *** (0.000) (0.000) (0.000) (0.000) Risks 2021, 9, 81 15 of 17 Table 8. Cont. Variables Interest Income to Gross Loans Ratio Model (1) Model (2) Model (3) Model (4) Bank industry concentration 0.016 *** 0.016 *** 0.019 *** 0.019 *** (0.000) (0.000) (0.000) (0.000) GDP growth rate 0.117 *** 0.088 *** 0.120 *** 0.108 *** (0.000) (0.000) (0.000) (0.000) Inflation 0.061 *** 0.066 *** 0.056 *** 0.061 *** (0.006) (0.003) (0.005) (0.002) Developing countries dummy 1.232 *** 1.323 *** 1.142 *** 1.171 *** (0.000) (0.000) (0.000) (0.000) Crises dummy 0.787 *** 0.983 *** 0.312 *** 0.379 *** (0.003) (0.000) (0.000) (0.000) Year FE Yes Yes No No Constant 7.385 *** 7.018 *** 7.798 *** 7.906 *** (0.000) (0.000) (0.000) (0.000) Observations 20,149 20,149 20,149 20,149 R-squared 0.820 0.819 0.819 0.818 There are two important points to consider while interpreting these results. First, the WUI and EPU measure different aspects of uncertainty. The EPU measures economic policy uncertainty while the WUI index represents overall economic uncertainty and is broader in definition. Usually, governments and central banks consider global trends to set their fiscal and monetary policies. On the other hand, country-specific economic events that create uncertainty are more local in nature. Second, such time dependence of the EPU index is less likely to affect the results of those empirical studies that control their models with time fixed-effects. 5. Conclusions In this paper, we examine the impact of economic uncertainty on bank loan pricing. We represent aggregate country-level economic uncertainty with the world uncertainty index (WUI) of Ahir et al. (2018). Using a cross-country bank-level dataset from 88 countries over the period 1998–2017, we find that banks increase interest rates on loans in response to heightened economic uncertainty. Specifically, a one-standard-deviation increase in WUI leads to a 20.67-basis-points increase in interest-income-to-gross-loans ratio. Our results are robust when we use the alternative proxy of uncertainty, include additional controls in our model, and use alternative estimation methods. We further find that the WUI index is better in measuring local economic uncertainty as compared to the EPU index. Our findings imply that policy uncertainty is an important risk factor in bank loan pricing and increases funding cost for firms and households. Policy uncertainty might increase in response to a crisis (i.e., financial, banking, or the pandemic as was the case with the outbreak of Covid-19 in early 2020) when a government has to choose among various possible potential policies. Though we control for the effect of crises in our regressions, future research may consider how policy uncertainty and crises interact to affect bank loan interest rates. Funding: No funding was obtained for this study. Institutional Review Board Statement: Not applicable. Informed Consent Statement: Not applicable. Data Availability Statement: Not applicable. Conflicts of Interest: The author declare no conflicts of interest. Risks 2021, 9, 81 16 of 17 Data Sharing: The data that support the findings of this study was collected from the Bankfocus database. Bankfocus is a proprietary database and doesn’t allow sharing data publicly. However, dataset used for this study can be requested from the corresponding author for research purposes. References Aaberge, Rolf, Kai Liu, and Yu Zhu. 2017. Political uncertainty and household savings. Journal of Comparative Economics 45: 154–70. [CrossRef] Ahir, Hites, Nicholas Bloom, and Davide Furceri. 2018. The World Uncertainty Index. Available online: https://ssrn.com/abstract=32 75033 (accessed on 10 August 2019). Ashraf, Badar Nadeem. 2017. Political Institutions and Bank Risk-Taking Behavior. Journal of Financial Stability 29: 13–35. [CrossRef] Ashraf, Badar Nadeem. 2020. Policy Uncertainty and Bank Liquidity Hoarding: International Evidence. Available online: http: //ssrn.com/abstract=3574193 (accessed on 19 April 2021). Ashraf, Badar Nadeem, and Yinjie Shen. 2019. Economic policy uncertainty and banks’ loan pricing. Journal of Financial Stability 44: 100695. [CrossRef] Asquith, Paul, Anne Beatty, and Joseph Weber. 2005. Performance pricing in bank debt contracts. Journal of Accounting and Economics 40: 101–28. [CrossRef] Azzimonti, Marina. 2018. Partisan conflict and private investment. Journal of Monetary Economics 93: 114–31. [CrossRef] Baker, Scott R., Nicholas Bloom, and Steven J. Davis. 2016. Measuring economic policy uncertainty. Quarterly Journal of Economics 131: 1593–636. [CrossRef] Becker, Ralf, Denise R. Osborn, and Dilem Yildirim. 2012. A threshold cointegration analysis of interest rate pass-through to UK mortgage rates. Economic Modelling 29: 2504–13. [CrossRef] Belkhir, Mohamed, Narjess Boubakri, and Jocelyn Grira. 2017. Political risk and the cost of capital in the MENA region. Emerging Markets Review 33: 155–72. [CrossRef] Berger, Allen N., Nathan H. Miller, Mitchell A. Petersen, Raghuram G. Rajan, and Jeremy C. Stein. 2005. Does function follow organizational form? Evidence from the lending practices of large and small banks. Journal of Financial Economics 76: 237–69. [CrossRef] Berger, Allen N., Omrane Guedhami, Hugh Hoikwang Kim, and Xinming Li. 2020. Economic policy uncertainty and bank liquidity hoarding. Journal of Financial Intermediation, 100893. [CrossRef] Bloom, Nicholas. 2009. The impact of uncertainty shocks. Econometrica 77: 623–85. Bloom, Nicholas. 2014. Fluctuations in Uncertainty. Journal of Economic Perspectives 28: 153–76. [CrossRef] Bloom, Nicholas, Max Floetotto, Nir Jaimovich, Itay Saporta-Eksten, and Stephen J. Terry. 2018. Really Uncertain Business Cycles. Econometrica 86: 1031–65. [CrossRef] Blot, Christophe, and Fabien Labondance. 2013. Business lending rate pass-through in the Enrozone: Monetary policy transmission before and after the financial crash. Economics Bulletin 2: 973–85. Bonaime, Alice, Huseyin Gulen, and Mihai Ion. 2018. Does policy uncertainty affect mergers and acquisitions? Journal of Financial Economics 129: 531–58. [CrossRef] Bordo, Michael D., John V. Duca, and Christoffer Koch. 2016. Economic policy uncertainty and the credit channel: Aggregate and bank level U.S. evidence over several decades. Journal of Financial Stability 26: 90–106. [CrossRef] Bradley, Daniel, Christos Pantzalis, and Xiaojing Yuan. 2016. Policy risk, corporate political strategies, and the cost of debt. Journal of Corporate Finance 40: 254–75. [CrossRef] Brand, Thomas, Marlène Isoré, and Fabien Tripier. 2019. Uncertainty shocks and firm creation: Search and monitoring in the credit market. Journal of Economic Dynamics and Control 99: 19–53. [CrossRef] Brogaard, Jonathan, and Andrew Detzel. 2015. The Asset-Pricing Implications of Government Economic Policy Uncertainty. Manage- ment Science 61: 3–18. [CrossRef] Chen, Yunsen, Deqiu Chen, Weimin Wang, and Dengjin Zheng. 2018. Political uncertainty and firms’ information environment: Evidence from China. Journal of Accounting and Public Policy 37: 39–64. [CrossRef] Cheng, Chak Hung Jack, Ching-Wai Jeremy Chiu, William B. Hankins, and Anna-Leigh Stone. 2018. Partisan Conflict, Policy Uncertainty and Aggregate Corporate Cash Holdings. Journal of Macroeconomics 58: 78–90. [CrossRef] Espinosa-Vega, Marco A., and Alessandro Rebucci. 2004. Retail bank interest rate pass-through: Is Chile atypical? Central banking, analysis, and economic policies book series. Banking Market Structure and Monetary Policy 7: 147–82. Francis, Bill B., Iftekhar Hasan, and Yun Zhu. 2014. Political uncertainty and bank loan contracting. Journal of Empirical Finance 29: 281–86. [CrossRef] Ge, Wenxia, Tony Kang, Gerald J. Lobo, and Byron Y. Song. 2017. Investment decisions and bank loan contracting. Asian Review of Accounting 25: 262–87. [CrossRef] Giavazzi, Francesco, and Michael McMahon. 2012. Policy uncertainty and household savings. Review of Economics and Statistics 94: 517–31. [CrossRef] Grechyna, D. 2018. Firm size, bank size, and financial development. Journal of Economic Dynamics and Control 97: 19–37. [CrossRef] Greenwald, Bruce C., and Joseph E. Stiglitz. 1990. Macroeconomic models with equity and credit rationing. In Asymmetric Information, Corporate Finance, and Investment. Chicago: University of Chicago Press, pp. 15–42. Risks 2021, 9, 81 17 of 17 Greenwald, Bruce C., Joseph E. Stiglitz, and Andrew Weiss. 1984. Informational imperfections in the capital market and macro- economic fluctuations. National Bureau of Economic Research. Available online: https://www.nber.org/system/files/working_ papers/w1335/w1335.pdf (accessed on 10 August 2019). Gregor, Jir ˇí, and Martin Melecký. 2018. The pass-through of monetary policy rate to lending rates: The role of macro-financial factors. Economic Modelling 73: 71–88. [CrossRef] Gulen, Huseyin, and Mihai Ion. 2016. Policy uncertainty and corporate investment. Review of Financial Studies 29: 523–64. [CrossRef] Houston, Joel F., Chen Lin, Ping Lin, and Yue Ma. 2010. Creditor rights, information sharing, and bank risk taking. Journal of Financial Economics 96: 485–512. [CrossRef] Hu, Shiwei, and Di Gong. 2019. Economic policy uncertainty, prudential regulation and bank lending. Finance Research Letters 29: 373–78. [CrossRef] Huang, Henry He, Gerald J. Lobo, Chong Wang, and Jian Zhou. 2018. Do Banks Price Independent Directors Attention? Journal of Financial and Quantitative Analysis 53: 1755–80. [CrossRef] Kanagaretnam, Kiridaran, Chee Yeow Lim, and Gerald J. Lobo. 2014. Influence of National Culture on Accounting Conservatism and Risk-Taking in the Banking Industry. Accounting Review 89: 1115–49. [CrossRef] Laeven, Luc, and Fabian Valencia. 2018. Systemic Banking Crises Revisited. IMF Working Paper. WP/18/206. Washington: International Monetary Fund. Li, Xiang, Bibo Liu, and Xuan Tian. 2018. Policy Uncertainty and Household Credit Access: Evidence from Peer-to-Peer Crowdfunding. Available online: https://ssrn.com/abstract=3084388 (accessed on 10 August 2019). Nguyen, Canh Phuc, Thai-Ha Le, and Thanh Dinh Su. 2020. Economic policy uncertainty and credit growth: Evidence from a global sample. Research in International Business and Finance 51: 101118. [CrossRef] Park, Jin Seok, and Donghyun Suh. 2019. Uncertainty and household portfolio choice: Evidence from South Korea. Economics Letters 180: 21–24. [CrossRef] Pastor, Lubos, and Pietro Veronesi. 2013. Political uncertainty and risk premia. Journal of Financial Economics 110: 520–45. [CrossRef] Pham, Anh Viet. 2019. Political risk and cost of equity: The mediating role of political connections. Journal of Corporate Finance 56: 64–87. [CrossRef] Phan, Hieu V., Nam H. Nguyen, and Hien T. Nguyen. 2019. Policy uncertainty and firm cash holdings. Journal of Business Research 95: 71–82. [CrossRef] Qian, Jun, and Philip E. Strahan. 2007. How laws and institutions shape financial contracts: The case of bank loans. The Journal of Finance 62: 2803–34. [CrossRef] Stiglitz, Joseph E., and Andrew Weiss. 1981. Credit Rationing in Markets with Imperfect Information. The American Economic Review 71: 393–410. Valta, Philip. 2012. Competition and the cost of debt. Journal of Financial Economics 105: 661–82. [CrossRef] Waisman, Maya. 2013. Product market competition and the cost of bank loans: Evidence from state antitakeover laws. Journal of Banking & Finance 37: 4721–37. Waisman, Maya, Pengfei Ye, and Yun Zhu. 2015. The effect of political uncertainty on the cost of corporate debt. Journal of Financial Stability 16: 106–17. [CrossRef] Yung, Kenneth, and Andrew Root. 2019. Policy uncertainty and earnings management: International evidence. Journal of Business Research 100: 255–67. [CrossRef] Zeng, Jianyu, Teng Zhong, and Fan He. 2019. Economic policy uncertainty and corporate inventory holdings: Evidence from China. Accounting and Finance 60: 1727–57. [CrossRef]

Journal

Risks – Multidisciplinary Digital Publishing Institute

Published: Apr 23, 2021

Keywords: bank loan interest rates; economic uncertainty; political uncertainty; world uncertainty index; economic policy uncertainty

There are no references for this article.