Liquidity dynamics of banks in emerging market economies

Mashamba, Tafirei Article Liquidity dynamics of banks in emerging market economies Journal of Central Banking Theory and Practice Provided in Cooperation with: Central Bank of Montenegro, Podgorica Suggested Citation: Mashamba, Tafirei (2022) : Liquidity dynamics of banks in emerging market economies, Journal of Central Banking Theory and Practice, ISSN 2336-9205, Sciendo, Warsaw, Vol. 11, Iss. 1, pp. 179-206, https://doi.org/10.2478/jcbtp-2022-0008 This Version is available at: https://hdl.handle.net/10419/299036 Standard-Nutzungsbedingungen: Die Dokumente auf EconStor dürfen zu eigenen wissenschaftlichen Zwecken und zum Privatgebrauch gespeichert und kopiert werden. Sie dürfen die Dokumente nicht für öffentliche oder kommerzielle Zwecke vervielfältigen, öffentlich ausstellen, öffentlich zugänglich machen, vertreiben oder anderweitig nutzen. Sofern die Verfasser die Dokumente unter Open-Content-Lizenzen (insbesondere CC-Lizenzen) zur Verfügung gestellt haben sollten, gelten abweichend von diesen Nutzungsbedingungen die in der dort genannten Lizenz gewährten Nutzungsrechte. Terms of use: Documents in EconStor may be saved and copied for your personal and scholarly purposes. You are not to copy documents for public or commercial purposes, to exhibit the documents publicly, to make them publicly available on the internet, or to distribute or otherwise use the documents in public. If the documents have been made available under an Open Content Licence (especially Creative Commons Licences), you may exercise further usage rights as specified in the indicated licence. https://creativecommons.org/licenses/by-nc-nd/4.0/ Liquidity Dynamics of Banks in Emerging Market Economies 179 * Great Zimbabwe University, Masvingo, Zimbabwe and University of South Africa, Pretoria, South Africa E-mail: [email protected] Journal of Central Banking Theory and Practice, 2022, 1, pp. 179-206 Received: 09 September 2020; accepted: 07 December 2020 UDK: 339.721:336.71 DOI: 10.2478/jcbtp-2022-0008 Tafirei Mashamba * Liquidity Dynamics of Banks in Emerging Market Economies Abstract: This study examines the liquidity dynamics of banks in emerging market economies. Using annual data of 91 commercial banks from 11 countries, the study established that banks in emerging markets have target liquidity ratios they pursue and partially adjust due to market frictions. Overall, risk aversion and prudence play a significant role in explaining the liquidity dynamics by banks in emerging market economies. Keywords: bank liquidity, liquidity dynamics, commercial banks, emerging markets, GMM. JEL Classification: G11, G18, G19, G21, G28. 1. Introduction Liquidity is of vital importance to banking institutions. On an ongoing basis, a bank has to ensure that it keeps ample cash and a stock of liquid securities to meet its contractual obligations such as cash withdrawals (Subramoniam, 2018, Casu, Girardone and Moluneux, 2006). Three issues are central to bank liquidity. First, the trade-off between liquidity and profitability. Banks keep liquidity buffers to mitigate liquidity risk; however, maintaining high levels of liquidity to mitigate liquidity risk has an opportunity cost in the form of interest income forgone by holding zero or low yield earning liquid assets. Second, banks’ balance sheets are fragile by construct which makes them susceptible to failure (Diamond and Dybvig, 1983). Third, bank liquidity problems are contagious due to the interconnectedness of banks and other financial intermediaries. Liquidity problems at an individual bank, especially a systemically important one, can quickly transcend to other banks and the real economy if it is not swiftly addressed (Van Journal of Central Banking Theory and Practice 180 Rixtel and Gasperini, 2013). Therefore, the significance of liquidity falls beyond an individual bank because idiosyncratic liquidity challenges can quickly spill over to other banks and financial institutions as well as the real economy. Besides, a lack of liquidity can be detrimental even to banks that are highly capitalised as revealed by events that transpired during the 2007/9 global financial crisis. A bank may be well-capitalised and profitable, but a loss of creditors’ confidence in the institution’s ability to settle obligations upon request may lead to sudden large “en-masse” withdrawals which may bring down an otherwise solvent institution (Bindseil and Fotia, 2021; Elliot, 2014). For instance, the Basel Committee on Banking Supervision (2013) and Le Lesle (2012) observed that although most banks entered the 2007/9 financial crisis with favourable capital ratios, liquidity shortages ignited and catalysed their failure. Accenture (2015) adds that banks did not develop proper liquidity projection models and they over-relied on volatile short term wholesale funding such as Repurchase Agreements (Repos) and Asset-Backed Commercial Paper (ABCP) to finance their activities. At the same time, banks invested heavily in structured products such as Asset-Backed Securities (ABS), which are vulnerable to illiquidity in times of severe financial stress such as the 2007/2009 financial turmoil (Caverzasi, Botta and Capelli, 2019; Kowalik, 2013). Virtually, all financial transactions and commitments affect a bank’s liquidity position. Moreover, a bank’s cash inflows and cash outflows are stochastic as they depend on market conditions and other agents’ behaviour (Basel Committee on Banking Supervision, 2008). This suggests that liquidity management in banking firms is a complex task: it requires bank managers to develop liquidity optimisation models to optimise their liquid assets holdings. Financial innovation and market dynamics have also brought changes in the ways that banks manage their liquidity. Traditionally, banks relied on retail deposits for funding. However, financial innovation has enabled banks to use short term debt instruments like commercial paper and repurchase agreements to source liquidity from the liability side of their balance sheets (Subhanij, 2010). Nevertheless, events that transpired during the 2007/9 mayhem caused banks to re-examine their liquidity management practices. This study attempts to shed some insights into the liquidity dynamics of banks in emerging markets given the importance of banks in these markets and the significance of emerging markets in the global economy. Most emerging market economies are bank-based (Tuna and Almahadin, 2021). This emanates from rudimentary and/or less developed capital markets. The study extends literature in the following ways. First, the research explores bank liquidity dynamics from Liquidity Dynamics of Banks in Emerging Market Economies 181 the asset side of the banks’ balance sheet (statement of comprehensive income). This motivation stems from the influence of market liquidity (asset sales) on a banks’ overall liquidity profile. Elliot (2014) posits that a bank’s liquidity position is significantly influenced by its ability to generate liquidity through asset sales which is dependent on market conditions. Second, as far as could be ascertained, this is the first study to empirically estimate a partial adjustment model for bank liquidity in emerging markets that determines the speed of adjustment. This approach is commonly used in corporate finance and capital management studies. Third, the study proffers insights into the strategic behaviour of liquidity management of banks in emerging markets. Lastly, but not least, the study is premised on commercial banks operating in emerging market economies. This scope is based on the intuition that liquidity management practices of banks are likely to vary between bank-based (emerging economies) and market-based (developed) economies due to differences in market structures and development. Yet, most empirical studies on bank liquidity management are drawn from advanced economies (for example, Banerjee and Mio; 2017; DeYoung and Jang, 2016; Bonner and Eijffinger, 2012). This study seeks to fill this gap. The rest of the study is organised as follows. The succeeding section discusses the variables that influence bank liquidity and formulates hypotheses; the third section attends to methods of the study; the fourth section explores the data (descriptive statistics), while section five presents and discuss the empirical results, with the last section looking at policy implications and recommendations. 2. Literature review: Factors affecting bank liquidity and hypotheses formulation Past levels of liquidity (LaRic, t-1 ) Studies by Mashamba and Kwenda (2017), DeYoung and Jang (2016) and Delechat, Arbelaez, Muthoora and Vtyurina (2012) show that banks’ liquidity ratios are persistent. Hence, as suggested by Louzis and Vouldis (2015), if the current values of a particular variable are influenced by its past values, the appropriate methodology for regression analysis is a dynamic error component panel model (partial adjustment model) that captures persistence in the dependent variable. For this reason, the study included the lagged dependent variable among the set of the explanatory variable to account for persistence in liquidity ratios and formulates the first hypothesis as follows: H1: Adjustment costs may influence banks to maintain liquidity buffers. Journal of Central Banking Theory and Practice 182 Bank capital (CAP) Two competing theories attempt to explain the relationship between bank capital and liquidity, namely financial fragility and risk absorption theory. The risk absorption theory is based on the literature of Repullo (2004), and Von Thadden (2004). Repullo and Von Thadden argue that since capital absorbs losses, it increases the bank’s capacity to bear risk which entices it to create more liquidity (by lending); therefore, banks with high levels of capital may target low liquidity. In addition, Bonner and Hilbers (2015) argue that adequately capitalized banks have better access to funding markets, due to their perceived low default risk; hence, they can operate with low levels of liquid assets. On the other hand, the financial fragility theory postulated by Diamond and Rajan (2000) predicts a positive relationship between bank capital and liquidity. Their argument is based on the intuition that bank capital may inhibit liquidity transformation (lending) since it makes a bank’s capital structure to be fragile. From this discussion, the relationship between bank capital and liquidity is ambiguous; therefore, the study expects either a positive or negative coefficient term. H2a: Bank capital positively influences bank liquidity adjustment. H2b: Bank capital negatively affects bank liquidity dynamics. Bank Size (SIZE) The “too big to fail” theory states that regulators are unlikely to permit large banks to fail out of the fear that their closure would trigger the widespread failure of other banks (Anginer, Demirgüç-Kunt, Huizinga and Ma, 2018). Consequently, large banks may target low liquidity on the belief that they will be bailed out. Moreover, large banks are characterised by stable cash flows, better access to capital markets, investment opportunities, and business diversification and their loan portfolios are highly likely to contain liquid assets like syndicated loans (DeYoung and Jang, 2016; Kochubey and Kowalczyk, 2014). In addition, big banks tend to command a large market share and market power (Gautam, 2016). Therefore, large banks have strong incentives to carry low levels of liquid assets. Thus, size is hypothesised to inversely affect bank liquidity. H3: Large banks have great incentives to target low levels of liquidity. Loan growth (LG) Lending is the principal business activity of commercial banks. As such, the amount of liquid assets maintained by a bank is significantly influenced by loan Liquidity Dynamics of Banks in Emerging Market Economies 183 demand (Alger and Alger, 1999). If loan demand is weak (strong), banks tend to hold more (less) liquid assets. The study, therefore, predicts that loan growth negatively affects bank liquidity. H4: Loan growth negatively affects bank liquidity adjustment. Asset quality (LLOSS) Based on the asset quality signalling hypothesis proposed by Lucas and McDonald (1992), asset quality determine bank liquidity adjustment dynamics (Kola, Gjipali and Sula, 2019). Loan loss reserves indicate the perceived riskiness of a bank’s loan portfolio. Lucas and McDonald (1992) argue that an increase in loan loss reserves is interpreted as a sign of potential distress by investors, which leads to reduced funding. This means that banks experiencing asset quality deterioration may suffer a significant decrease in external liquidity support. Similarly, Tabak, Li, Vasconcelos and Cajueiro (2013) assert that a rise in loan defaults decreases the amount of liquidity that a bank can generate from loan repayments. Thus, banks expecting high loan losses should maintain high levels of liquidity to ameliorate liquidity risk. Apriori, the study expects a positive association between loan loss provisions and banks’ liquid assets holdings. H5: Loan-loss provisioning positively affects bank liquidity adjustment. Profitability (ROE) Profits represent a ready source of liquidity to a bank since huge business profits improve a firm’s cash holdings which in turn boost its liquidity (Aspachs, Nier and Tiesset, 2005). This implies that profitable banks may hold significant amounts of liquidity. On the contrary, Bonner and Eijffinger (2012) contend that profitability reduces banks' incentives to maintain large liquidity buffers. They argue that profitable banks can easily fund themselves with debt, due to their ability to service debts, when confronted with liquidity shocks, which makes them be less liquidity constrained. Based on these arguments, the relationship between profitability and banks' liquid assets holdings is ambiguous; hence, the study expects either a positive or negative coefficient term. H6: A significant rise in profits enables banks to easily adjust their liquidity levels (H6a: β6>0). However, huge profits can create incentives for banks to target lower liquidity due to an increased ability to use capital markets for funding (H6b: β6<0). Journal of Central Banking Theory and Practice 184 Deposit-loan synergy (DLS) Banks offer liquidity services to both depositors and borrowers by offering checking accounts to depositors and loan commitments (credit lines) to borrowers. In the course of providing these services, banks expose themselves to liquidity risk. Banks can hedge this risk by combining transaction/demand deposits and loan demand (Kashyap, Rajan and Stein, 2002). As long as cash demand from depositors is uncorrelated with credit line draw-downs by borrowers, banks can use cash inflows from demand deposits to satisfy loan commitment requests, thereby enabling them to reduce cash holdings while serving both clients (Gatev, Schuermann and Strahan, 2007). This strategy is known as the deposit-loan synergy, and it reduces a bank’s impetus to maintain large liquidity buffers for precautionary reasons. H7: Deposit-loan synergy reduces banks’ incentives to maintain large liquidity buffers. Transaction deposits (TD) One of the primary roles of commercial banks in an economy is to offer maturity transformation services to economic agents, that is, to accept short term deposits and issue long term loans. Consequently, the principal source of liquidity to commercial banks tends to be transaction (demand) deposits (Singh and Sharma, 2016). As such, banks with high levels of demand deposits are expected to be highly liquid. Likewise, given that withdrawal of transaction deposits is unpredictable, demand deposits carry a high risk of unexpected withdrawals; hence, as transaction deposits increase, banks should invest more in liquid assets to ameliorate liquidity risk (Chen and Phuong, 2014). The study, therefore, predicts that banks with large transaction deposits target low liquidity. H8: Banks with large transaction deposits target low liquidity. Deposit Insurance (DEP) Besides bank-specific characteristics discussed above, the study also considered deposit insurance to be a significant factor that explains bank liquidity holdings. The presence of deposit insurance removes incentives for depositors to run on an institution thereby reducing the bank’s liquidity risk and ultimately its liquidity buffers (Diamond and Dybvig, 1983). Thus, banks operating in countries with explicit deposit insurance schemes may be less worried about “en masse” withdrawals or bank runs; hence, they may target low liquidity buffers. Apriori, the study predicts an inverse relationship between deposit insurance and banks' li- Liquidity Dynamics of Banks in Emerging Market Economies 185 quidity buffers. Deposit insurance is captured by a dummy variable (DEPINS) that equals one for a country with deposit insurance coverage and zero otherwise. Data on countries' deposit insurance status were obtained from a comprehensive database on deposit insurance schemes created by Demirgüç-Kunt, Kane and Laeven (2014) at the end of 2013. H9: The presence of a deposit insurance scheme removes incentives for banks to target large liquidity buffers. Business Cycles (GDP) In a world characterised by capital market imperfections, banks’ liquidity buffers tend to be countercyclical (Aspachs et al., 2005; Delechat et al., 2012). Countercyclicality refers to a scenario whereby banks accumulate liquidity reserves (hoard liquidity) in times of weak economic prospects due to high default risk and weak loan demand and draw down their buffers (lend) in times of economic booms, in response to increased lending opportunities and low default risk. Accordingly, this study hypothesises that business cycles negatively influence banks’ liquidity buffers. The study uses annual growth in the real gross domestic product (GDP) as a proxy for business cycles. H10: Banks react to economic booms by lending aggressively, thereby targeting lower liquidity. (H10a: γ1 >0). Conversely, when the economy moves into a recession banks respond to the economic meltdown by hoarding liquidity (H10b: γ1<0). Savings (SR) In general, corporate and household savings find their way to banks either through direct deposits or investments in banks’ debt products (Pati and Shome, 2011). As such, banks operating in countries with a high level of savings should be associated with high levels of bank liquidity (Wadesango, Lora and Charity, 2017). Therefore, the study expects savings to positively influence bank liquidity. H11: Savings positively influence bank liquidity adjustments. Monetary Policy (CBR) In many jurisdictions, central banks attempt to influence economic activity using various tools, especially short term interest rates (the central bank rate or policy rate). Their intervention is likely to affect banks’ liquidity adjustments since monetary policy is transmitted via banks (Awdeh, 2019). When the central bank cuts (hike) interest rates, banks tend to respond to this policy change by maintaining Journal of Central Banking Theory and Practice 186 few (large) amounts of liquid securities relative to total assets (Aspachs et al., 2005). Stated differently, monetary policy tightening tends to be associated with low liquid assets holdings while monetary policy loosening results in increased liquid assets holdings by commercial banks. Therefore, this study hypothesises that bank liquidity is negatively related to policy rates. H12: Bank liquidity adversely responds to policy rates. 3. Methodology 3.1. Sample and data This study is based on a representative sample of commercial banks operating in eleven emerging market economies, namely Hong Kong, India, Mexico, Saudi Arabia, South Africa, Argentina, Indonesia, Korea, Russia, Singapore, and Turkey. The sample is made up of ninety-one (91) banks. The number of banks from each economy is presented in Appendix I. The study period is confined to the period January 2011 to December 2016 which is post the global financial crisis and pre-COVID 19 pandemics. This period was chosen because it eliminates structural breaks that are associated with the global financial crisis and the Covid19 pandemic. The data for individual banks were obtained from Income Statements and Balance Sheets. The data were retrieved from the Bankscope Bureau Van Dijk database. Macroeconomic data for each respective country were obtained from the World Bank databank. 3.2. Empirical model and estimation approach Liquidity management at banking institutions can be examined in the context of the trade-off theory which is mainly used in corporate finance studies. The theory states that firms target an optimal amount of liquid securities that balance the benefits and costs of maintaining liquid assets (De Haan and Hinloopen, 2003; Kim, 1998). The benefits of holding liquid assets are two-fold: transaction and speculative purposes. The transaction motive suggests that firms maintain liquidity buffers to avoid transaction costs that are related to sourcing external funding and the need to liquidate assets to pay off maturing liabilities. The speculative motive submits that firms keep liquid assets to exploit new investment opportunities that may arise since external funding may not be available as and when needed or costly. On the other hand, the costs associated with liquidity buffers are interest income that is foregone as a result of investing in low yield Liquidity Dynamics of Banks in Emerging Market Economies 193 Table 2: Results of banks liquidity management practices Variable Model 1 Model 2 Coefficient Sign (1) Economic impact (2) Coefficient Sign (3) Economic impact (4) LARic,t-1 0.5467*** (0.1508) -0.6681*** (0.1212) - SIZE 5.8783** (2.9607) 0.8702 2.0368 (2.6470) 0.3015 CAP - 0.0917 (0.2373) -0.0181 -0.2147 (0.2667) -0.0423 LG 0.0513*** (0.0148) 0.0555 0.0899*** (0.0168) 0.0971 LLOSS -2.283*** (0.5783) -0.2022 -1.8096*** (0.5237) -0.1603 ROE -0.1947*** (0.0286) -0.1074 -0.1382*** (0.0294) -0.0762 DLS -0.2321*** (0.0390) -0.2681 -0.2014*** (0.0435) -0.2327 TD 11.9923** (5.6294) 0.7941 12.7741* (7.7582) 0.8459 DEPINS 63.4001 (97.4963) 0.8022 9.9682 (96.7466) 0.1261 GDP 1.8842** (0.8626) 0.1812 1.4419 (0.9295) 0.1387 SR -1.3611*** (0.4114) -0.3995 -1.7993*** (0.5144) -0.3400 CBR -0.4843 (0.5559) 0.0598 -0.8904 (0.5925) - 0 .110 0 Time fixed effects No No Yes Yes Arellano-Bond (2) test Sargan test Wald test 0.6190 0.5911 914.68*** 0.6273 0.4704 2516.42*** Source: Own construction based on data from Bankscope. ***, **, * denotes 1%, 5% and 10% significance level respectively. Standard errors in the parenthesis (brackets). Lagged liquidity ratio (LaRt-1) The coefficient of the lagged dependent variable is positive and statistically significant at 1% significance level. Therefore, the adoption of a dynamic panel model in this study is substantiated. The positive and significant coefficient of the lagged dependent variable suggests that banks in emerging market economies have target liquidity levels and they partially adjust their liquidity to reach their desired Journal of Central Banking Theory and Practice 194 liquidity level consistent with the trade-off theory. Moreover, this evidence suggests that liquidity ratios banks in emerging market economies are persistent and banks in emerging markets actively managed their liquidity over the period of study. This finding is consistent with Delechat et al. (2012) finding that liquidity ratios of banks in Central America are persistent. Without time dummies, the speed at which banks adjust their liquidity to revert to their target level is estimated to be 0.4533 (1-0.5467). These results imply that banks close about 45% of deviation from their desired liquidity level within a year. At this speed of adjustment, it would take roughly 2.21 years to reach their target. After controlling for time fixed effects, the speed at which banks in emerging market economies adjust their liquidity decreases to 0.3319 (i.e. 1-0.6681). The speed at which banks in emerging market economies adjust their liquidity is slow. This slow adjustment speed is consistent with the proposition that adjustment costs preclude banks to immediately revert to their target liquidity level, thereby confirming the hypothesis that adjustment costs create incentives for banks to maintain liquidity buffers (H1). As discussed earlier some of the factors that influence adjustment costs are market frictions such as asymmetric information, transaction costs, and agency costs. These market frictions create strong incentives for banks to minimize adjustment costs by holding higher levels of liquidity. This evidence concurs with Drobetz, Schilling and Schroder (2014) finding that adjustment costs tend to be high in bank-based (emerging) economies relative to market-based (developed) economies because advanced economies have welldeveloped and vibrant capital markets which make it relatively easy for banks to adjust their liquidity. As evidence, Ernst and Young (2013) reports that stock market capitalization as a proportion of GDP is about four times higher in advanced economies compared to emerging markets economies. Ernst and Young went on to add that developed economies bond market size is almost 2.5 times greater than established emerging market economics like Malaysia and South Africa. Furthermore, a comparative analysis of adjustment speeds of banks in bankbased economies and market-based economies may offer additional evidence to this analysis. In the United States of America, De Young and Jang (2016) found that banks in the United States of America adjust their liquidity by approximately 27.15% per annum, meaning that they close 27% of the gap between their target and desired liquidity in a year. Their results demonstrate that banks in the United States of America target and actively manage their liquidity. The findings of De Young and Jang and the present study’s empirical results support the proposition that adjustment costs are higher in bank-based economies relative to market- Liquidity Dynamics of Banks in Emerging Market Economies 195 based economies. Consequently, difficulties in assessing external funding may explain why banks in emerging market economies hold excess liquidity. Bank capital (CAP) The coefficient of parameter is statistically insignificant in both models; thus, H2 could not be verified by empirical results. These findings imply that capital has no significant impact on the size of the liquidity buffer maintained by banks in emerging market economies. One plausible explanation to these findings could be that although capital creates incentives for banks to keep low liquidity, its impact could have been affected by Basel III capital requirements. Basel III package requires banks to maintain both large liquidity and capital ratios. The joint management of liquidity and capital requirements might have reduced the influence of capital on banks’ liquidity adjustments. Bank Size (SIZE) The point estimate is positive and very significantly different from zero (5.8783) in Model 1. In terms of economic impact, the coefficient elasticity evaluated at the sample mean is 0.8702. This means a one standard deviation change in bank size contributes to 87% changes in bank liquid assets holdings, indicating that size significantly explains the size of liquidity buffers maintained by banks. This evidence refutes the hypothesis that big banks maintain low levels of liquidity (H3) and lend support to the conjecture that small banks depend more on themselves in liquidity management by keeping large liquidity buffers probably because they have limited access to external funding. These results concur with the findings of (Lastuvkova, 2014) who examined liquidity management strategies of banks in the Czech Republic and found that small banks invest more in liquid assets compared to large banks, for precautionary reasons as they have limited external financing. Loan growth (LG) Empirical results show that the relationship between bank liquidity and loan growth is positive and statistically significant in both models. In terms of economic significance, a one standard deviation change in loan growth leads to about a 6% increase in bank liquidity (Model 1 results, Column 2, Table 2). Contrary to empirical evidence from developed markets, for example, Kochubey and Kowalczyk (2014); the study could not find evidence at conventional levels to support the conjecture that banks in emerging markets experiencing high loan growth maintain low liquidity. Journal of Central Banking Theory and Practice 196 Asset quality (LLOSS) Contrary to expectations, the point estimate of loan loss reserves to gross loans is negative with a coefficient of -2.283 and it is statistically significant at 1% significance level in the absence of time dummies. Hence, the study could not find evidence to support the claim that banks in emerging markets respond to asset quality deterioration by increasing their liquid asset holdings in anticipation of reduced external funding. This evidence lends supports to the principal argument of this study that banks in emerging markets depend less on capital markets for funding or banks in emerging economies rely more on themselves (deposits) for funding. Profitability (ROE) The coefficient of ROE is negative and significantly different from zero (-0.1947) at 1% level. The elasticity of bank profitability computed at the sample mean is -0.1074 (Table 2, Model 1, Column 1,). A 19.47% increase in bank profits triggers banks in emerging markets to reduce their liquid assets holdings by roughly 11%, all things constant. It appears profitable banks in emerging market economies are less financially constrained, implying that they can easily raise external funding when the need arises. This decreases their need to maintain large liquidity reserves. Stated differently, empirical results suggest that profitable banks tend to maintain low liquidity because they experience less financial constraints when borrowing from funding markets, possibly because they can service debts. This finding concurs with Delechat et al. (2012) finding that profitable banks in Central America tend to keep low liquidity because they can easily obtain external funding from capital markets when they face liquidity shocks. Moreover, these results are consistent with empirical evidence from advanced economies. For example, Bonner and Eijffinger's (2012) study found that profitable banks in the Netherlands operate with low levels of liquidity because they can easily access funding from capital markets as they have ample cash to service their debts. Deposit – Loan Synergy (DLS) The point estimate of the variable DLS has a negative sign -0.2321(Column 1, Table 2) and is statistically significant at 1% level in the baseline model. In terms of economic significance, a one standard deviation increase in deposit-loan synergy practice triggers banks in emerging markets to reduce liquid assets holdings by roughly 27%, ceteris paribus. The empirical results substantiate the proposition that commercial banks hedge liquidity risk through deposit-loan synergies (H7) Liquidity Dynamics of Banks in Emerging Market Economies 197 consistent with evidence from advanced economies Kashyap et al. (2002) and Gatev et al. (2007). Transaction Deposits (TD) The point estimate of the variable transaction deposits is positive and statistically significant (11.9923) at 5% level in the baseline model (model without time dummies). Its elasticity computed at sample mean in the baseline model is 0.7941 (Table 2, Column 2). When transaction deposits increase by about 12 units banks’ investments in liquid securities grow by about 0.79 units, all else equal. These findings suggest that bank deposits and liquidity increase (decrease) jointly in emerging markets, supporting the claim that banks with large demand deposits tend to pursue large liquidity buffers (H8). This evidence is consistent with empirical findings from developed economies (for instance, De Haan and Van den End, 2013). Deposit insurance (DEPINS) Contrary to expectations, empirical results show that the point estimate of deposit insurance on banks' liquidity is positive, but not significantly different from zero (the p-value is 0.5162 in Model 1). Therefore, the hypothesis that deposit insurance coverage incentivises banks in emerging market economies to keep low levels of liquidity is not confirmed. Business cycles (GDP) The coefficient of real GDP growth is positive and statistically significant at 1% significance level. The estimated coefficient of 1.8842 in the model without time dummies corresponds to a sensitivity value of 0.1812 (Column 2, Table 2). A oneunit increase in real GDP growth contributes to a 0.1812 unit increase in bank liquidity, all things constant. The positive association between business cycles and bank liquidity implies that bank liquidity in emerging market economies is procyclical (Kozarić and Žunić Dželihodžić, 2020). This evidence conveys that banks in emerging market economies build up their liquidity holdings when the economy is doing well and run down their buffers when the economy enters into a recession. Therefore, the study found evidence to confirm H10. 2 Not reported for brevity. Journal of Central Banking Theory and Practice 198 Savings level (SR) Surprisingly, empirical results indicate that savings negatively affect bank liquidity. A 1.36 unit increase in savings motivates banks to decrease investments in liquid assets by about 0.40 units, all else equal. Consequently, (H11) could not be supported. These results could imply that banks in emerging market economies invest less national savings in liquid securities, however, it seems they channel most of the savings towards productive investments aimed at spurring economic growth and job creation. This evidence may render support to the notion that firms in emerging market economies mainly rely on banks for long term funding since banks appear to be investing most of their savings deposits in loans. This view is in line with the Financial Stability Board's (2011) assertion that emerging markets are characterised by concentrated and less complex financial systems and banks play a large role in financial intermediation because capital markets and other financial institutions are still underdeveloped. Monetary policy (CBR) The estimated coefficient of the central bank rate is negative in both models but statistically insignificant. Consistent with the International Monetary Fund (2009), the study could not find enough statistical evidence at conventional levels to support the hypothesis that monetary policy affects banks' liquidity adjustments in emerging markets. The International Monetary Fund suggests that the ineffectiveness of monetary policy in emerging markets may be attributed to global financial crisis strains that might have buckled monetary policy transmission in emerging market economies. Furthermore, the insignificant interplay between monetary policy and banks’ liquidity buffers could be attributed to high liquidity reserves maintained by banks in emerging markets that makes monetary policy ineffective. 6. Policy Implications and Recommendations This study was interested in providing insights into liquidity adjustment dynamics and management techniques pursued by banks in emerging markets economies. Research findings revealed that banks in emerging market economies have target/optimal liquidity levels and they partially adjust to maintain their desired liquidity level. The speed of adjustment was found to be slow suggesting that banks in emerging economies face high adjustment costs. In light of these findings, it can be inferred that adjustment costs create incentives for banks in emerging markets to maintain liquidity buffers. Liquidity Dynamics of Banks in Emerging Market Economies 199 Furthermore, the study established that bank-specific characteristics influence liquidity adjustment decisions of banks in emerging markets. The finding that bank size positively influences banks’ liquidity adjustment implies that banks in emerging market economies depend more on liquid assets and less on wholesale funding for liquidity management. From this evidence, the study can conclude that banks in emerging markets are risk-averse. In terms of policy implications, this behaviour engenders banking sector stability; hence, policymakers should reinforce it through strict monitoring of banks’ compliance with the liquidity coverage ratio (LCR) regulation. Moreover, the study established that banks in emerging markets increase liquid assets holdings as their lending business grows. Since maturity transformation exposes banks to liquidity risk, empirical results suggest that banks in emerging markets are risk-averse as they increase holdings of liquid assets in response to the growth in loans (illiquid assets). Moreover, this behaviour demonstrates prudent liquid management. From these results, it can be inferred that banks in emerging markets conservatively and prudently manage their liquidity. Regulators in emerging markets ought to reinforce this good practice by monitoring the compliance of banks to the liquidity coverage ratio (LCR) rule which encourages banks to maintain liquid assets that correspond to their expected net cash outflows over 30 days. Research findings also revealed that banks in emerging markets with large volumes of transaction deposits maintain large liquidity buffers, suggesting that banks in emerging markets react to growing transaction deposits by increasing investments in liquid assets. This practice demonstrates sound liquid management; hence, regulators should strengthen this good behaviour through strict supervision of the LCR standard. Another interesting finding worth mentioning is the negative impact of loan loss reserves ratio on banks’ liquid assets adjustment. This finding suggests that banks in emerging market economies poorly manage credit risk and this has some implications for both bank managers and supervisors. Loan loss provisions are important because they play a significant role in determining the stability and soundness of banking institutions. Inadequate loan loss provisioning may result in capital erosion which jeopardises the banking sector’s stability. As such, banks’ loan loss provision estimates are a vital tool for microprudential regulation that regulators use to monitor the quality of banks’ loan portfolios. Based on these empirical findings, bank managers in emerging markets should adopt forwardlooking loan loss management practices. Such practices are consistent with IFRS 9 impairment rules. Likewise, due to asymmetric information between regula- Journal of Central Banking Theory and Practice 200 tors and banks, bank regulators need to obtain timely information on banks’ loan loss provisions since loan losses are reported on an accrual basis. Delays in obtaining such information in time would paint a good picture of banks’ solvency which may not be true. This evidence reinforces the introduction of IFRS 9 in banking institutions. The study also contributes to the analysis of the relationship between macroeconomic conditions and banks’ liquidity holdings. The positive association between real GDP growth and banks’ liquidity buffers suggests that bank liquidity is procyclical, meaning that banks in emerging markets accumulate (drawdown) liquidity buffers when the economy is performing well (badly). This behaviour is consistent with the aims of the LCR. The LCR encourages banks to build up liquidity buffers in good times and draw them down in terms of crisis. As such, the study advocates policymakers to reinforce this interplay through tight supervision of liquidity requirements. Monetary policy in emerging market economies was found to be ineffective in altering overall banking sector liquidity in emerging markets. This implies that central banks’ efforts to stimulate economic activity by reducing short term interest rates are defeated in emerging markets probably because aggregate bank liquidity is very high. This calls for regulators to look at other tools of stimulating economic activity by manipulating bank liquidity. One such tool is statutory reserve requirements. This analysis possibly explains why statutory reserve requirements are very high in emerging markets. Overall, the study established that risk aversion and prudence play a significant role in explaining the level of liquid assets maintained by banks in emerging market economies. Liquidity Dynamics of Banks in Emerging Market Economies 201 References 1. Accenture, (2015). Liquidity Coverage Ratio: Implications and a Pragmatic Approach to Implementation. Accenture. Available at: https://www. accenture.com/_acnmedia/Accenture/ConversionAssets/DotCom/ Documents/Global/PDF/Dualpub_9/Accenture-Liquidity-Coverage-Ratio. pdf (Assessed 26 January 2019) 2. Alger, G. and Alger, I. (1999). Liquid Assets in Banks: Theory and Practice, Boston College Department of Economics, Boston (Working Paper No. 446). 3. Anginer, D., Demirgüç-Kunt, A., Huizinga, H. and Ma, K. (2018). Corporate governance of banks and financial stability.Journal of Financial Economics,130(2): 327-346. 4. Angklomkliew, S., George, J. and Packer, F. (2009). Issues and developments in loan loss provisioning: the case of Asia. Bank for International Settlements Quarterly Review, December: 69-93. 5. Aspachs, O., Nier, E, and Tiesset, M. (2005). Liquidity, Banking Regulation and the Macroeconomy. Evidence on bank liquidity holdings from a panel of UK-resident banks. Unpublished manuscript. Bank for International Settlements: Basel. 6. Awdeh, A., (2019). Monetary policy and economic growth in Lebanon.Journal of Central Banking Theory and Practice,8(2): 147-171. 7. Basel Committee on Banking Supervision, (2008). Principles for Sound Liquidity Risk Management and Supervision. Bank for International Settlements, Basel, Switzerland. 8. Basel Committee on Banking Supervision, (2013). Basel III: The Liquidity Coverage Ratio and liquidity risk monitoring tools. Bank for International Setttlements, Basel, Switzerland. 9. Banerjee, R. N. and Mio, H. (2017). The impact of liquidity regulation on banks. Journal of Financial Intermediation, 11(7): 1-15. 10. Basso, O., Delgado, F.L. and Meza, M. (2012).Strengthening Bank Capital and Liquidity in Central America: The Road to Basel III. International Monetary Fund Mimeo, Washington DC, United States of America. 11. Bindseil, U. and Fotia, A. (2021). Financial Instability. InIntroduction to Central Banking(pp. 67-78). Springer, Cham. 12. Blundell, R. and Bond, S. (1998). Initial conditions and moment restrictions in dynamic panel data models. Journal of econometrics, 87(1): 115-143. 13. Bonner, C. and Eijffinger, S.C.W. (2012). The Impact of the LCR on the Interbank Money Market, Discussion Paper 2012-075, Tilburg University, Center of Economic Research Journal of Central Banking Theory and Practice 202 14. Bonner, C. and Hilbers, P. (2015). Global liquidity regulation-Why did it take so long?, Dutch National Bank, Amsterdam. (Working Paper No. 455). 15. Caverzasi, E., Botta, A. and Capelli, C. (2019). Shadow banking and the financial side of financialisation.Cambridge Journal of Economics,43(4): 1029-1051. 16. Casu, B., Girardone, C., and Molyneux, P. (2006). Introduction to banking, New Jersey: Prentice Hall. 17. Chen, I-J. and Phuong, L.N. (2014). The determinants of Bank Liquidity Buffer. Mediacast. Available: https://mediacast.blob.core.windows.net/ production/Faculty/StoweConf/submissions/swfa2014_submission_9.pdf (Accessed on 14 May 2019). 18. De Haan, L. and Hinloopen, J., 2003. Preference hierarchies for internal finance, bank loans, bond, and share issues: evidence for Dutch firms.Journal of Empirical Finance,10(5): 661-681. 19. De Haan, L. and Van den End, J.W. (2013). Bank liquidity, the maturity ladder, and regulation. Journal of Banking and Finance, 37(10): 3930-3950. 20. Delechat, C., Arbelaez, C.H., Muthoora, P.S. and Vtyurina, S. (2012). The Determinants of Banks' Liquidity Buffers in Central America, International Monetary Fund, Washington DC, United States of America. (Working Paper No.12/301). 21. DeYoung, R. and Jang, K. Y. (2016). Do banks actively manage their liquidity? Journal of Banking and Finance, 66(C): 143-161. 22. Demirgüç-Kunt, A., Kane, E, and Laeven, L. (2014). Deposit Insurance Database. International Monetary Fund: Washington DC. 23. Diamond, D. W. and Dybvig, P.H. (1983). Bank runs, deposit insurance and liquidity. The Journal of Political Economy, 91(3): 401-419. 24. Diamond, D. W. and Rajan, R. G. (2000). A theory of bank capital. The Journal of Finance, 55(6): 2431-2465. 25. Drobetz, W., Schilling, D.C. and Schroder, H. (2014). Heterogeneity in the Speed of Adjustment across Countries and over the Business Cycle, Available: http://citeseerx.ist.psu.edu/viewdoc/download?doi=10.1.1.641.2420andrep=r ep1andtype=pdf (Accessed on 30 July 2019). 26. Elliott, D. J. (2014). Bank Liquidity Requirements: An Introduction and Overview. The Brookings Institution. Available: https://www.brookings. edu/wpcontent/uploads/2016/06/23_bank_liquidity_requirements_intro_ overview_elliott.pdf (Accessed on 5 February 2019). 27. Ernst and Young (2013). Banking in emerging markets Seizing opportunities, overcoming challenges. London, United Kingdom. 28. Financial Stability Board, (2011). Financial Stability Issues in Emerging Market and Developing Economies. Report to the G-20 Finance Ministers and Central Bank Governors. Available: https://openknowledge.worldbank.