Liquidity and Deposit Money Banks’ Performance in Nigeria

[Pages:7]Dynamic Research Journals (DRJ) Journal of Economics and Finance (DRJ-JEF) Volume 2 ~ Issue 4 (April, 2017) pp: 55-61 ISSN (Online); 2520-7490

Liquidity and Deposit Money Banks' Performance in Nigeria

1Agu Nkechi Christiana & 2Dr. Emeka Nnamani

1Department of Banking and Finance, Enugu State University of Science and Technology (ESUT), Email: nkechiucheagu@. 2Department of Business Administration, Enugu State University of Science and Technology (ESUT), Email: emekannamani98@ .

Abstract: The study investigated the relationship between liquidity and performance of Deposit Money Banks in Nigeria. Specifically, the study investigated the effect of loan-to-deposit ratio and interest rate on return on equity of DMBs in Nigeria. Loan-to-deposit ratio and interest rate served as the explanatory variables. On the other hand, return on equity served as the dependent variable and served as a measure of performance in the DMBs. The study covered the period 2005 to 2015 and data were collected from the Annual reports of First Bank Plc and Fidelity Bank Plc. The study employed the Pooled (Panel) Least Squares (PLS) technique in order to investigate the `aggregate effect' of liquidity on the selected Deposit Money Banks (DMBs) in Nigeria. Findings showed that liquidity has a negative and insignificant relationship with DMBs performance in Nigeria while interest rate has a positive and significant effect on DMBs performance in Nigeria. The study recommends that policy makers should formulate and implement policies that would curb unnecessary risk taking in the banking industry so as to prevent the banks from overshooting that threshold at which the loan-to-deposit ratio increases rather than decrease banks' performance. Keywords: Loan-to-deposit ratio, Deposit Money Banks, Liquidity and Return on equity

I. Introduction

Most often than not, the influence of liquidity on the performance of banks is undermined. This stems from the notion held by the traditional asset pricing models which asserts that the financial market is without frictions and therefore liquidity is assumed not play any roles at all (Adler, 2012). In the context of the traditional asset pricing model, the financial intermediaries were assumed to be stable because they had access to readily available funds and they obtain these funds at low cost. This notion was popular among financial analysts prior to the outbreak of the Global Financial Crisis (GFC). However, the importance of liquidity in the banking industry came to the fore with the happenings since the outbreak of the global financial crisis of 2008/2009. The banks became endangered species as they were faced with huge distress simply because they undermined the importance of liquidity to the system and therefore did not manage it efficiently (Tarraf & Majeske, 2014). It cannot go without noticing that the rapid downward trend in market conditions helped to explain how quickly liquidity can disappear. More shocking, as evidenced by the Global Financial Crisis, is the fact that illiquidity can drain already earned profits. This is because the financial institutions are either forced to sell accumulated assets below their market value or they are forced to borrow at interest rates that are well above their return on assets (performance). By and large, what began as credit challenges for the United States' sub-prime market alone virally exploded into challenges in global credit markets with huge losses (Shen, Chen, Kao and Yeh, 2009). As a result of volatility experienced in the global credit markets, financial market players became increasingly risk-averse and became unwilling to invest in any market other than the ones where their returns are guaranteed. Their actions led to reduced levels of liquidity in the global financial markets and adversely affected the banks' performances (SARB, 2009).

Back home in Nigeria, the then Nigerian Governor of Central Bank assured Nigerians that the Nigerian banks would not be affected even as liquidity challenges ravage other world financial institutions. His assurances came on the heels of the robustness of Nigerian banks which he hinged on the much touted bank consolidation (recapitalization) exercise of year 2005. Is the performance of Nigerian banks actually not affected by liquidity challenges? This is the pertinent question that motivated the study.

1.1 Statement of the Problem Given that financial intermediation by the banks (through loan disbursement and deposit mobilization) increases economic activities and enhances economic growth thereby increasing banks' performance, the challenge remains where the banks will strike a balance as to the threshold which their loans shall not exceed. This is against the



55 | P a g e

Liquidity and Deposit Money Banks' Performance in Nigeria

backdrop that at a certain threshold the higher the loan-to-deposit ratio the higher the performance of the banks but beyond this threshold high loan-to-deposit ratio adversely affects liquidity thereby undermining bank performance.

1.2 Research Objectives

The broad objective of the study was to examine the relationship between liquidity and Deposit Money Banks'

performance in Nigeria. Specifically, the study did the following:

(i)

Investigated impact of loan-to-deposit ratio on return on equity of Deposit Money Banks (DMBs) in

Nigeria.

(ii) Investigated the impact of interest rate on return on equity of Deposit Money Banks (DMBs) in

Nigeria.

1.3 Research Questions

In line with the specific objectives, the following questions sought to be answered by the study:

(i)

To what extent does loan-to-deposit ratio affect return on equity of Deposit Money Banks (DMBs) in

Nigeria?

(ii) To what extent does interest rate affect return on equity of Deposit Money Banks (DMBs) in Nigeria?

1.4 Research Hypotheses

Two hypotheses were tested in the study and they are:

(i)

H0: Loan-to-deposit ratio does not have significant effect on return on equity of Deposit Money Banks

(DMBs) in Nigeria.

(ii) H0: Interest rate does not have significant effect on return on equity of Deposit Money Banks (DMBs)

in Nigeria.

II. Empirical Literature

Kosmidou, Tanna and Pasiouras (2005) investigated the relationship between liquidity and bank performance in United Kingdom for the period 1995-2002. The study adopted return on assets as the measure for banks' performance while ratio of liquid assets to customer and short term borrowings served as the measure for liquidity. Findings revealed that ratio of liquid assets to customer and short term borrowings have a positive and significant effect on banks' performance in Greece. Similarly, Kosmidou (2008) in an independent study investigated the determinants of bank performance in Greece using a sample of 23 Greek banks. This study adopted data from 1990 to 2002 and employed the Pooled Ordinary Least Squares (POLS) as the analytical tool. Findings revealed that a liquidity was a major determinant of bank performance in Greece and that liquidity has a negative relationship with performance of Greece banks as less liquid banks have lower return on assets (proxy for performance). Naceur and Kandil (2009) examined the effect of capital requirements on banks' cost of intermediation and performance in Egypt. The study employed capital-to-assets ratio, management efficiency, and increase in liquidity and inflation rate as the independent variables while both return on assets and return on equity served as the dependent variables. Findings revealed that liquidity does not have any significant effect on both return on assets and return on equity.

Chen, Shen, Kao and Yeh (2010) investigated the effect of liquidity on bank performance. The study employed net interest margin as the dependent variable while liquidity risk served as the independent variable. Their findings revealed that a positive and significant relationship exists between liquidity risk and the performance of banks suggesting that banks with higher levels of illiquid assets achieve higher interest incomes. The study further revealed that liquidity risk has a negative significant relationship with return on average assets and return on average equity in a market-based financial system. However, the study revealed that there is no significant relationship between liquidity risk and bank performance in a bank-based financial system since the banks play a key role in providing financing. Thus, they are not affected by liquidity risk in a bank-based financial system. Marozva (2015) analyzed the relationship between liquidity and bank performance in South Africa for the period 1998 to 2014. The study employed the autoregressive distributed lag (ARDL) model and the Ordinary Least Squares (OLS) model as the analytical tools. Unit root and cointegration pre-testing were conducted first to determine the stationarity and existence of long run equilibrium relationship among the variables. The study categorized market liquidity risk, funding liquidity risk and credit liquidity risk and used them as the key measures for liquidity while net interest margin served as the measure for performance. Findings showed that there exists a negative significant relationship between net interest margin and funding liquidity risk but an insignificant relationship exists between market liquidity risk and credit liquidity risk; and net interest margin (performance) in South African Banks.



56 | P a g e

Liquidity and Deposit Money Banks' Performance in Nigeria

Sujeewa (2015) investigated the impact of credit risk management on the performance of commercial banks in Sri Lanka. The study employed both descriptive and empirical tools as analytical methodologies. Specifically, the study employed the Panel data analysis making use of data from 2009 to 2013. Return on assets served as the measure of performance as well as the dependent variable while non-performing loans to total loans ratio, loan provision to non-performing loan ratio, loan provision to total assets ratio and loan provision to total loans ratio served as the measures of credit risk as well as the independent variables. Findings revealed that non-performing loans and provisions have an adverse effect on commercial banks' performance in Sri Lanka. Dhanuskodi (2014) analyzed the effect of loan deposit ratio on profitability of banks in Malaysia. The study adopted data from 2009 to 2013 and employed a Panel data approach to empirically investigate the effect of loan-to-deposit ratio on return on assets. The study revealed that there exist a positive and insignificant relationship between loan-to-deposit ratio and profitability of commercial banks in Malaysia. Olarewaju and Adeyemi investigated the relationship between liquidity and profitability of Nigeria Deposit Money Banks. The study employed the causality test and findings showed that there exists a unidirectional relationship between liquidity and profitability in the DMBs with the flow of direction of influence going from liquidity to profitability. The implication of this finding is that liquidity is a strong determinant of the profitability of the DMBs in Nigeria.

2.1 Theoretical Framework

Some theories of liquidity ? commercial banks nexus were reviewed in the study. Among them are:

2.1.1 Commercial Loan Theory This theory is also referred to as traditional theory of liquidity and it was popularized by Onoh (2002). This

theory postulates that lending by the banks should mainly be on short term basis since most bank deposits are also in short term. Hence, the theory emphasizes the need to march short term deposits with short term loans in order for the banks to be more efficient in liquidity management.

2.1.2 Anticipated Loan Theory This theory was postulated by Herbert Prochnow in 1949 and emphasizes that the earning power and credit

worthiness of the borrower should be the top consideration of the banks when granting loans as it was the major source of bank liquidity. Hence, it becomes imperative for the banks to critically examine the reputation of the borrower, his ability and willingness to repay the borrowed funds as and when due. If the borrowers have the ability to service and /or pay back as required, the profitability (performance) of the banks would be enhanced. This theory favours this study as emphasis is on the approach to adopt so as to eliminate the rising incidence of non-performing loans.

2.1.3 Liquid Asset Theory This theory was popularized by Anyanwu (1993) and postulates that banks should maintain large pool of

short term assets. With efficient primary and secondary markets, such short term assets would be able to facilitate the banks' abilities to meet its short term obligations as the fall due.

III. Model Specification

The study anchored on the anticipated loan theory which emphasizes on the need for the banks to identify credible and willing borrowers so as to reduce the incidences of non-performing loans and enhance performance of the DMBs. In line with Marozva (2015) with modifications, the model for the study is specified as:

ROE = (LDR, INTR) ..................... (1) Transforming equation (1) into its linear econometric form yields

ROE = 0 + 1LDR + 2INTR + .................. (2) Where; ROE = Return on equity (proxy for DMBs performance), LDR = Loan-to-deposit ratio (proxy for funding liquidity), INTR = Interest rate (used as a control variable), 0 = Constant term, 1 and 2 = Coefficient parameters of the explanatory variables, = Stochastic error term. By a priori, 0 > 0, 1 < 0, and 2 < 0

3.1 Description of Variables

(i) Dependent variable Return on Equity (ROE); Return on equity measures the performance of the DMBs. Return on equity is described in the study as: ROE = Profit after Tax divided by Shareholders' equity



57 | P a g e

Liquidity and Deposit Money Banks' Performance in Nigeria

(ii) Independent Variables Loan-to-deposit ratio (LDR); Loan-to-deposit ratio is used as a measure for liquidity in the DMBs. It is measured as the ratio of total loans to total deposits. Loan-to-deposit ratio is described in the study as: TL/TD = Total loans divided by total deposits.

Interest rate (INTR); Interest rate is used as a measure for monetary policy effect. It is captured as monetary policy rate as stipulated by the Central Bank of Nigeria and used merely as a control variable in the study. It is described in the study as: INTR = Monetary Policy rate (MPR)

IV. Data Analysis and Discussion of Findings

Table 1: Data on Return on equity (ROE), Loan-to-deposit ratio (LDR) and Interest rate (INTR)

YEAR

BANK

ROE

LDR

INTR

2005

FIRSTBANK

27.3

55.59

15

2006

FIRSTBANK

27.2

44.94

13

2007

FIRSTBANK

23.7

37.67

12.25

2008

FIRSTBANK

9

66.17

8.75

2009

FIRSTBANK

0.4

63.83

9.81

2010

FIRSTBANK

7.9

76.45

7.44

2011

FIRSTBANK

17.6

74.45

6.13

2012

FIRSTBANK

17.9

75.13

12

2013

FIRSTBANK

16.9

71.35

12

2014

FIRSTBANK

18.8

79.25

12.25

2015

FIRSTBANK

18.9

65.09

14

2005

FIDELITY

12.72

76.2

15

2006

FIDELITY

12.35

58.99

13

2007

FIDELITY

13.98

43.58

12.25

2008

FIDELITY

9.56

63.12

8.75

2009

FIDELITY

1.78

60.4

9.81

2010

FIDELITY

4.33

62.03

7.44

2011

FIDELITY

0.81

47.45

6.13

2012

FIDELITY

11.1

48.2

12

2013

FIDELITY

4.72

52.84

12

2014

FIDELITY

7.97

66.06

12.25

2015

FIDELITY

7.58

75.13

14

Source: Author's compilation from selected banks' annual reports and Central Bank of

Nigeria (CBN) Statistical Bulletin.

4.1.1 Return on equity Table 1 above shows that there are different levels of performance attached to different banks at different

times (periods). For instance, the performance of First Bank Plc stood at 27.3 percent in 2005 and this decreased to 7.9 percent in 2010. However, it resumed its upward trajectory reaching 18.9 percent in 2015. On the other hand, the performance of Fidelity stood at 12.72 percent and continued to increase reaching 13.98 percent in 2007. Thereafter, the performance of Fidelity bank was characterized by rapid increases and decreases in the remaining years. Therefore, it is pertinent to re-emphasize that the selected banks have different levels of performance for the years under review. This may be a pointer to the fact that the banks may be differently affected by liquidity. In order to bring to the fore the `sum-effect' of liquidity on the banks, the study employed the Pooled Ordinary Least Squares technique.



58 | P a g e

Liquidity and Deposit Money Banks' Performance in Nigeria

4.1.2 Loan-to-deposit ratio (LDR) The effect of liquidity risk on the performance of the DMBs was captured using the loan-to-deposit ratio.

From table 1 above, one can see that the loan-to-deposit ratio differs among the banks at different periods. For instance, in 2005, the loan-to-deposit ratio of First bank Plc stood at 55.59 percent while that of Fidelity bank Plc stood at 76.2 percent. In year 2010, the loan-to-deposit ratio of First bank Plc increased to 76.45 percent while that of Fidelity bank Plc decreased to 62.03 percent. These dynamics in the loan-to-deposit ratio in the banks continued to play out thereby indicating that the banks undertake different levels of risk in terms of what the loan out in relation to the bank deposits.

4.1.3 Interest rate (INTR) Interest rate was adopted as a control variable in the study. It represented the external environment in which

the banks operate in and to that extent it is assumed that its effect is `central' to the banks. For instance, in 2005, the monetary policy rate (interest rate) stood at 15 percent and this decreased to 7.44 percent in 2010 but increased again to 14 percent in 2015. Being that the effect is central; the effect on the banks is expected to be the same. Table 2: Pooled Ordinary Least Squares (POLS) Result Method: Panel Least Squares Date: 03/01/17 Time: 06:04 Sample: 2005 2015 Periods included: 11 Cross-sections included: 2 Total panel (balanced) observations: 22

Variable

Coefficient Std. Error t-Statistic Prob.

C LDR INTR

-0.510211 10.64673 -0.060680 0.128475 1.494283 0.577748

-0.047922 -0.472313 2.586393

0.9623 0.6421 0.0181

R-squared Adjusted R-squared S.E. of regression Sum squared resid Log likelihood F-statistic Prob(F-statistic)

0.571557 0.494878 7.145443 970.0898 -72.86646 3.541508 0.049290

Mean dependent var S.D. dependent var Akaike info criterion Schwarz criterion Hannan-Quinn criter. Durbin-Watson stat

12.38636 7.963398 6.896951 7.045729 6.931998 1.817524

Critical values: (a) t statistic, t0.05 = 1.721 (b) F-statistic, F0.05 (2, 19) = 3.47

Source: Author's computation using E-views 8.0 software The result in table 2 above can be summarized as: ROE = -0.51 ? 0.06LDR + 1.49INTR t-statistic (-0.05) (-0.47) (2.59) R-squared = 0.57 F-statistic = 3.54 DW-statistic = 1.82

The result summarized above is analyzed in line with the economic, statistical and econometric criteria. First, the result shows that there is a negative and insignificant relationship between loan-to-deposit (proxy for liquidity) and return on equity (proxy for performance) in DMBs in Nigeria. This result conforms to economic a priori expectation and shows that one unit increase in the loan-to-deposit ratio (liquidity risk) leads to 0.06 unit fall in the return on equity of Deposit Money Banks in Nigeria. The computed t-statistic for loan-to-deposit ratio (0.45) in absolute terms is less than the critical t-statistic (1.72) at five percent level of significance. As a confirmation, the



59 | P a g e

Liquidity and Deposit Money Banks' Performance in Nigeria

probability value of loan-to-deposit ratio (0.6421) is greater than the test significant level (i.e. P > 0.05). We conclude that loan-to-deposit ratio (liquidity risk) does not have significant effect on the performance of DMBs in Nigeria. This finding conforms to economic theoretical expectation to the extent that as banks increase their loan-todeposit ratio, the riskier their operations become and if the credit risk management is not effective their performance would be affected. Perhaps, this result is attributed to the increasing appetite of the banks to take uncertain risks which have resulted in rising incidence of non-performing loans in the Nigerian banking industry. As the nonperforming loans increase, the performance of the banks reduces. As non-performing loans rise, liquidity squeeze increases and the banks become unable to obtain sufficient fund either by increasing liabilities or converting its assets promptly at a reasonable cost. All these ultimately undermine the performance of the bank. Despite the fact that an increasing loan-to-deposit to a certain threshold may enhance the performance of banks, but beyond that threshold an increase in loan-to-deposit ratio reduces the performance of the banks. This finding corroborates Naceur and Kandil (2009) which established a negative and insignificant relationship between liquidity and financial performance of banks.

Second, the result reveals that there is a negative and significant relationship between interest rate and return on equity of Deposit Money Banks in Nigeria. By sign, this result conforms to economic a priori expectation because an increase in interest rate charged on loans leads to a discouragement among borrowers thereby reducing the performance of the banks. From the result, one unit rise in interest rate leads to 1.49 unit fall in the return on equity (performance) of the Deposit Money Banks in Nigeria. However, the computed t-statistic for interest rate (2.59) is less than the critical (tabulated) t-statistic (1.72) at five percent level of significance. As a confirmation, the probability value of interest rate (0.0181) is less than the test significant level (i.e. P < 0.05). Hence, we conclude that interest rate has an insignificant effect on the performance of Deposit Money Banks (DMBs) in Nigeria. This relationship is not in line with economic expectation because it is expected that as interest rate increases, the cost of borrowing increases leading to many borrowers becoming discouraged from borrowing. With a reduction in the number of borrowers, the returns those transactions would have brought to the banks are reduced thereby reducing the performance of the banks. This finding concurs with the findings of Okoye and Eze (2013) which argued that a positive and significant relationship exists between monetary policy rate and performance of DMBs in Nigeria. Perhaps, this outcome may be attributed to dearth of borrowing alternatives in Nigeria which makes the borrowers to still approach the banks irrespective of the unfavourable interest rate regimes of the banks. As this trend continues despite the high interest rate, the banks rake in more profits and their level of performance increases.

The coefficient of determination (R?squared) shows that 57 percent of the variations in Deposit Money Banks' performance in Nigeria are due to changes in loan-to-deposit ratio and interest rate. Therefore, the remaining 43 percent are caused by changes in other factors not included in the model. The computed F-statistic (3.54) exceeds the tabulated F-statistic (3.47) and this shows that the model adopted for the study is significant as well as reliable. The Durbin ? Watson statistic (1.82) lies within the permissible region and indicates that there is no autocorrelation. Interestingly, the Durbin ? Watson statistic is greater than the R ? squared and this confirms that the regression result is not spurious.

V. Summary of Findings

The findings of the study are summarized below:

(i)

There exists a negative relationship between loan-to-deposit ratio and performance of Deposit Money

Banks in Nigeria.

(ii) Loan-to-deposit ratio has an insignificant effect on the performance of Deposit Money Banks in

Nigeria.

(iii) There exists a positive relationship between interest rate and performance of Deposit Money Banks in

Nigeria.

(iv) Interest rate has a significant effect on the performance of Deposit Money Banks in Nigeria.

5.2 Conclusion The study undertook to examine the relationship between liquidity and Deposit Money Banks' performance in

Nigeria. To break the broad objectives into manageable specific objectives, the study investigated the effect of loanto-deposit ratio and interest rate on the return on equity. Loan-to-deposit ratio was used as a proxy for liquidity and interest rate was adopted merely as a control variable and both of them served as the independent variables. On the other hand, return on equity was used to measure performance of the DMBs and served as the dependent variable. Findings in the study showed that loan-to-deposit reduces the performance of DMBs in Nigeria but its effect on the banks is insignificant. Surprisingly, the study revealed that interest rate increases the performance of DMBs in Nigeria and has a significant effect on DMBs performance in Nigeria. In conclusion, the study asserts that although



60 | P a g e

Liquidity and Deposit Money Banks' Performance in Nigeria

liquidity reduces the performance of the DMBs but it does not have significant effect on their performance in Nigeria.

5.3 Recommendations

The study recommends the following based on its findings:

(i)

Policy makers in the banking industry should formulate and implement policies that would discourage

unnecessary risk-taking by the banks in order not to overshoot that threshold at which loan-to-deposit

ratio starts to have diminishing returns.

(ii) Deposit Money Banks in Nigeria should sustain programmes and packages that would further attract

borrowers such that even if the interest rate remains high it would not undermine their performance.

References [1]. Adler, D. (2012). The new field of liquidity and financial frictions. Research Foundation Literature Review, 7(2): 1-37. [2]. Ahmed, M. B. (2009). Measuring the performance of Islamic banks by adapting conventional ratios. German University in Cairo, Faculty of Management Technology Working Paper, 16: 1-26. [3]. Anyanwu, J. C. (1993). Monetary economic system in Nigeria. Lagos: Quodro Impressions Ltd. [4]. Chen, Y. K., Kao, L., Shen, C., & Yeh, C. (2010). Liquidity risk and its management in the Lithuanian banking system. Verslas XXI Amziuje Business in XXI Century, 6(1): 64-71. [5]. Dhanuskodi, R. (2014). Impact of loan deposit ratio (LDR) on profitability: Panel evidence from commercial banks in Malaysia. Proceedings of the 3rd International Conference on Global Business, Economics, Finance and Social Sciences, Mumbai, India, 19-21 December. [6]. Kosmidou, K. (2008). The determinants of banks' profit in Greece during the period of EU financial integration. Managerial Finance, 34: 146-159. [7]. Kosmidou, K., Tanna, S., & Pasiouras, F. (2005). Determinants of profitability of domestic UK commercial banks: Panel evidence from the period 1995-2002. Money, Macro and Finance (MMF) Research Group Conference. [8]. Marozva, G. (2015). Liquidity and bank performance. International Business and Economics Research Journal, 14(3): 453-461. [9]. Naceur, S. B., & Kandil, M. (2009). The impact of capital requirements on banks' cost of intermediation and finance: The case of Egypt. Journal of Economics and Business, 61: 70-89. [10].Okoye, V., & Eze, O. R. (2013). Effect of bank lending rate on the performance of Nigerian Deposit Money Banks. International Journal of Business and Management Review, 1(1): 34-43. [11].Olarewaju, O. M., & Adeyemi, O. K. (2015). Causal relationship between liquidity and profitability of Nigeria Deposit Money Banks. International Journal of Academic Research in Accounting, Finance and Management Sciences, 5(2): 165-171. [12].Onoh, J.K. (2002). Dynamics of money, banking and finance in Nigeria. An emerging market. Aba: Astra Meridian Publishers. [13].Prochnow, H.V. (1949). Term loan and theories of bank liquidity. New York: Prentice Hall, Inc. [14].Samad, A. (2004). Bahrain commercial banks' performance during 1994-2001. Credit and Financial Management Review, 10(1): 33-40. [15].SARB (2009). Financial stability Review, Pretoria: South African Reserve Bank. Available at http: //reservebank.co.za [16].Shen, C-H., Chen, Y-K., Kao, L-F., & Yeh, C-Y. (2009). Bank liquidity risk and performance. National Taiwan University Review, 3-37. [17].Tarraf, H., & Majeske, K. (2014). Impact of risk taking on bank financial performance during 2008 financial crisis. Oakland University.



61 | P a g e

 ................
................

In order to avoid copyright disputes, this page is only a partial summary.

Google Online Preview   Download