Financial Development and Economic Growth in Nepal: An ARDL Bound Test Approach

This study empirically examines the dynamic relationship between financial development and economic growth in Nepal using annual time series data from 1985 to 2016. The financial development is measured by domestic credit to the private sectors, domestic credit to the private sectors by banks, broad money (M2) and net domestic credit, separately. All are ratios to GDP. The economic growth is measured by real GDP per capita. The bound test approach of cointegration under autoregressive distributed lag (ARDL) model reveals that Nepal’s financial development and economic growth are cointegrated with bi-directional causality in the long-run. Thus, the study concludes that financial development and economic growth positively and significantly impact each other. The causal effects running from financial development to economic growth are portent then economic growth to financial development. However, the speed of adjustment towards long-run equilibrium, directing from economic growth to financial development is reasonably robust. There is one-directional reverse causality running from economic growth to financial development in the short-run. Therefore, the study suggests policymakers to prioritize policies to develop a well-functioning financial sector to enhance economic growth, especially for developing countries like Nepal. All measures of financial development are significant and positive to economic long


Introduction
The acquisition and interaction of production factors and technological transformations explain the countries' variances in economic growth and productivity. Besides this, financial intermediation is also evolving as a critical channel of economic growth and productivity in the present globalized world where labour and capital are moving across the countries rapidly (P. Demetriades and Law 2006). The financial sector plays an essential role in the economic and the prolonged political transition has deteriorated the development and diversification of the Nepalese financial sector. Therefore, examining the causal dynamics of financial development and economic growth of Nepal is very important for policymakers as Nepal has implemented the federal administration system after the promulgation of new constitutions in 2015.
Some existing studies have already examined the relationship between Nepal's financial development and economic growth (Bhetuwal 2007;Kharel andPokhrel 2012；Gautam 2014;Timsina 2014). However, most of these studies are either equipped with traditional econometric methods or motivated by 'finance leading growth' hypothesis. They do not consider their two-way causal dynamics for the long-run and short-run. Depending on these issues, this study attempts to enrich the available literature of the finance-growth relationship by providing new empirical evidence, evaluating a Nepalese perspective. For this, the study establishes two different hypotheses. The principle hypothesis assumes that financial development is the function of economic growth and reverse hypothesis assumes that economic growth is the function of financial development. The bound test approach of cointegration under the autoregressive distributed lag (ARDL) model examines the cointegration form and long-run dynamics. The error correction model (ECM) under ARDL is used to estimate short-run dynamics. Finally, one period lagged error correction term (ECT) confirmed the speed of adjustment towards long-run equilibrium.
The organization of this paper is as follows. Section 2 explains a summary of financial reform and development in Nepal. Section 3 briefly explains the literature review, section 4 presents data and proxy measures, and section 5 explains the econometric approaches and empirical models. Section 6 presents empirical analysis and discussions, section 7 highlights the conclusion and recommendation, and section 8 presents brief options for future research. Results obtained from these analyses might help to set clear policies in financial sector development, especially for developing countries like Nepal.

Financial Reform in Nepal
After facing an economic crisis in the early 1980s, Nepal introduced the first step of liberalization on the economy (Maskay and Subedi 2009)

. The government and International
Monetary Fund (IMF) made a Stand-By Agreement (SBA) in 1985, which introduced the first phase of financial reform under the Economic Stabilization Program (ESP) (Ozaki 2014). The program was intended to devaluate Nepalese currency, restrict public expenditure and bank credits, liberalize industrial licensing, promote export, and control import. The Structural Adjustment Program (SAP) of the World Bank in 1987 is implemented for further liberalization. Only two state-owned banks dominated the banking industry by holding 70% of the financial sector's total assets before implementing SAP (Ozaki 2014). Policies such as indirect monetary control, interest rate deregulation, the open market economy, liberal exchange rate, import licensing system, auction system of government securities made a fundamental change in financial system (P. O. Demetriades and Luintel 1996). The World Bank helped restructuring and strengthening of stateowned banks, amendment of the income tax act, commercial bank act, central bank act, and national industrial development corporation act and establishment of credit information bureau.
In 1999, IMF and the World Bank jointly conducted Financial Sector Assessment Program (FSAP) and stated that the Nepalese financial system was still fragile, vulnerable and risky concerning fundamental norms and principles of Basel Accords-1988 (Maskay and Subedi 2009  and reform financial institutions, to establish national banking training institute, to strengthen debt recovery tribunal, and to develop microfinance credit information services. As a result, the enhancement of supervision and monitoring strengths of the central bank with maintaining its independence and diversifying the financial sector in terms of size, ownership, operations, and investment has been achieved within a short period in Nepal.

Financial Development in Nepal
The establishment of the first commercial bank in 1937 for capital accumulation and promotion of trade and industry introduced the formal institutional development of Nepal's modern (SMEs) and credit expansions in highly potential sectors like tourism, green energy, and agroprocessing industries is still low in Nepal. Therefore, the central bank is enforcing commercial banks to expand their credit portfolio to the agricultural sector, energy sector and SME sectors by maintaining a minimum threshold of 15%, 10% and 15% of their total credit (NRB 2020). Nepal's central bank has set a clear vision of achieving sustainable and inclusive economic growth by "maintain macroeconomic and financial stability through proactive and effective monetary and financial policies" in their term in the third strategic plan 2017-2021. For this, the financial sector must mobilize available resources in the most productive and feasible sectors in the economy efficiently by increasing financial access and boosting the economy's financial stability.

Review of Literature
Historically, Schumpeter's (1911) influential studies and Hicks (1969) raised the financial sector's implication to encourage innovations through proper resource allocation in the economic development process. In contrast, Robinson (1952) and Goldsmith (1969) explained that economic development processes promote entrepreneurial activities and stimulate the financial sector. These two perspectives from the pioneer economists have raised a diverse opinion among researchers in finance-growth relationship literature. They have raised three debatable hypotheses about the finance-growth relationship. The most contradictory hypotheses are 'finance leading growth' hypothesis and 'growth leading finance' hypothesis. The third hypothesis asserts their mutual relationship of leading each other. Very few pieces of evidence show non-causality between them.
However, Patrick (1966) focused on both hypotheses and stated that the financial sector leads to economic growth in the preliminary phases of economic development and reverse in post-stage economic development. Shaw (1973) and McKinnon (1973) raised the implication of government regulations for repression and liberalization of the financial sector. They stated that financial liberalization enhances savings, encourage domestic investment, and boost economic growth.
Revisiting these policy arguments, Lucas (1988 p.6), stated that "the financial sector's role in economic growth is over-stressed" in previous studies. Chandavarkar (1992) raised neglected questions about the financial sector's needs and obstacles in developing countries concerning financial repression policies. King and Levine's (1993a) substantial effort confirms that the financial system promotes entrepreneurship by mobilizing savings in innovative and productive activities by diversifying risk and enhancing economic growth. Later on, King and Levine (1993b) supported Schumpeter's (1911) viewpoint empirically, arguing that financial sector development significantly determines the strength of accumulating capital, productivity, and efficiency of economic activities. The seminal works of Ross Levine and his co-authors with empirical pieces of evidence and broad literature stated that there exist cross-country variances on economic growth concerning size, depth, policies, and access of the financial sector (Levine 1997(Levine , 2005Levine and Zervos 1998).
However, some evidence shows that finance-growth relationship varies concerning economic development stages of the countries (Blackburn and Hung 1998;Christopoulos and Tsionas 2004;P. O. Demetriades and Hussein 1996;Hassan, Sanchez, and Yu 2011;Levine, Loayza, and Beck 2000;Minea and Villieu 2010). For example, Christopoulos and Tsionas (2004) claim that financial intermediation constitutes higher returns on economic activities in the middle stage of economic development. However, Hassan et al. (2011) said that economic growth gives more benefits to the financial sectors in the preliminary economic development phase, especially in vulnerable countries. Despite these diverse opinions, many researchers believe in the constructive linkages between financial intermediation and economic activities because they grow together, causing each other (Calderón and Liu 2003;Gregorio and Guidotti 1995;Jung 1986;Kar et al. 2011).
Few studies have examined the finance-growth relationship according to the firm's financial structure. Firms obtaining finance from external sources and optimizing their capital structure through financial intermediation, especially in countries with a more developed financial system, positively enhance productivity and growth (Arestis, Demetriades, and Luintel 2001;Rajan and Zingales 1996). The capital market based financial structure determines the sizeable economic growth for advanced economies but not for developing economies (Luintel et al. 2016). The capital market and credit market have an encouraging influence on real sector growth, but the credit market's contribution makes the more significant influence (Durusu-Ciftci, Ispir and Yetkiner 2016). However, an over-expansion of the financial sector or beyond optimum level might introduce volatility and diminish economic growth (Arcand, Berkes, andPanizza 2012, 2015;Law and Singh 2014;Beck, Degryse, and Kneer 2014;Samargandi, Fidrmuc, and Ghosh 2015).
Therefore, there is no clear consensus about the finance-growth relationship's causality directions (Martin Čihák et al. 2013). All countries may not benefit equally from the expansion of the financial sector. Instead, the benefits depend on regulatory and supervisory strengths and the effectiveness of the financial policies regarding services, stability, structure, and access (Barajas, Chami, and Yousefi 2013). Hence, the financial sector's success highly relies on its efficiency and effectiveness and the regulatory authorities' supervision and monitoring strength.
In Nepal's case,  studied the role of banking sector policies and their coexistence and estimated its effect on financial deepening. They further tested 'finance leading growth' hypothesis but did not consider the 'growth leading finance' hypothesis. G. K. Shrestha (2004) studied financial sector reform program of Nepal and emphasized on their proper implementation. M. B. Shrestha, and Chowdhury (2006) has constructed a financial liberalization index for Nepal but did not examine its impact on financial development. Bhetuwal (2007) has studied the causal relationship between financial liberalization index and various proxies of financial development. Kharel and Pokhrel (2012) have examined the causal relationship between Nepal's financial structure and economic growth and reveal that credit market promoted economic growth rather than the capital market in Nepal. Timsina (2014) has examined Nepal's economic growth effects through 'finance leading growth' hypothesis but missed to evaluate the reverse effects. Study of Gautam (2014) stated that reverse causality runs from economic growth to financial development in the long-run in Nepal, suggesting that further reform and expansion is needed for the efficiency and effectiveness of Nepal's financial sector. A recent study of Bist and Bista (2018) has tried to address the two-way dimensions of the finance-growth relationship addressing significant structural breaks of Nepal's financial development and economic growth.
However, their study depends on only one indicator of financial development.
Therefore, there is clear space to explore the two-way dynamics of the finance-growth relationship employing various financial development measures concerning the effect of Nepal's real sector and the external sector. Hence, this study tries to fulfil this gap by enriching the available literature with new empirical evidence using dynamic estimation methods in Nepalese context.

Data
This study's main objective is to examine the cointegrating association and causality dynamics between Nepal's financial development and economic growth for the policy recommendation. For this, the study examines annual data of 32 years from 1985 to 2016 obtained from world development indicators (WDI). The study periods cover an era of financial liberalization, policy reform and structural reform in the Nepalese economy. Since Nepal is still under financial liberalization process and should have a stable financial system to cope with internal and external shocks, we believe that obtained results in this study may have various advantages to formulate financial sector development policies for developing countries like Nepal. Table 1 presents the list of selected variables, their indications and short definitions. The graphical representation of the trend of these variables at their level values is presented in Appendix A.

Economic Growth
There are particular proxy measures of economic growth. This study measures economic growth, calculating annual changes in real GDP per capita (GDPPC). As Mankiw (1995) explained, the study follows the neoclassical growth model to measure economic growth. Thus, economic growth for the one year is defined as: Where, LnGDPPC is the natural logarithms of GDP per capita constant 2010 US dollar, and t represents the number of time-series observations.

Financial Development
Financial development is a multidimensional phenomenon concerning depth, access, efficiency and stability of financial system including financial institutions and markets of an economy (Almarzoqi, Ben Naceur, and Kotak 2015;Beck et al. 2008;M. Čihák et al. 2012).
However, the financial depth variables, such as narrow money and broad money supply (King andLevine 1993b, 1993a;Arestis and Demetriades 1997;Levine, Loayza, and Beck 2000;Kar, Nazlioǧlu, and Aǧir 2011) from monetary aggregates, total bank credit and deposit from financial institutions (Christopoulos and Tsionas 2004;Luintel and Khan 1999), and stock market capitalization and stock traded from financial markets (Arestis, Demetriades, and Luintel 2001;Levine and Zervos 1998), and their GDP ratios are most common among researchers. This study prefers to use a bank-based financial indicator and broad form of monetary aggregates following (Levine, Loayza, and Beck 2000;King andLevine 1993a, 1993b) to measure Nepal's financial development because Nepal has bank-based financial system rather than market-based. They are the domestic credit to the private sector (DCP), domestic credit to the private sector by banks (DCB), broad money (M2) as net liquid assets (BM) and net domestic credit (NDC). All of them are ratio to GDP. A high level of domestic credit to the private sector indicates higher credit access for the private sector, indicating the strength of capital formation of an economy. Levine (2005) stated that if a financial sector allocates more credit to the private sectors, they engage in researching firms, enhancing their corporate governance and control, and mobilizing savings by managing investment risks, and facilitating transactions. A high level of the domestic credit to the private sector by banks indicates the higher dependence of the private sector on the banking sector than the non-banking sector (Hassan, Sanchez, and Yu 2011). It indicates a higher level of financial development because of five special functions of banks suggested by Levine (1997). The higher level of the broad money (M2) measured as net liquid assets implies higher financial intensity and mobility and explains the strengths of channelling liquid funds from surplus sector to deficit sectors in an economy (Hassan, Sanchez, and Yu 2011;Khan and Senhadji 2003). A higher level of net domestic credit (NDC) indicates general credit mobilization's strengths to the government, nonfinancial public sector, and the economy's private sector.
The study has followed the neoclassical growth model to measure financial development, as Mankiw (1995) explained. Thus, financial development for the one year is defined as: Where, LnFD is the natural logarithms of four proxies of financial development, i.e. DCP, DCB, BM, NDC, used separately and t represents the number of time-series observations.

Other Control Variables
This study has used four additional variables to control the relationship between financial development and economic growth. They represent the magnitude of the real sector and the external sector of an economy. A steady economic growth heavily depends on domestic savings, which can be converted into investment activities through financial intermediation (Pagano 1993).
The financial intermediation plays a crucial role to channel those savings into investment. So, financial development and economic growth are expected to benefit from the level of gross domestic savings. So, this study uses gross domestic savings (GDS) ratio to GDP to control the estimations.
On the other hand, many countries are profoundly dependent on international trade to accelerate their economic activities, determining the real sector's magnitude. Nepal is heavily dependent on international trade. Inward remittance channelling from financial institutions is mostly used to finance imports because of low domestic production. Therefore, the merchandise trade as a sum of exports and imports ratio to GDP is also used as trade openness (TRD) to control the estimations.
Beside this, Nepal's frequently changing governments exercise budgetary spending through fiscal policies to influence the economy. The size of the fiscal budget and policies may affect financial and economic activities of an economy. Therefore, the general government final consumption expenditure ratio to GDP as the size of government (GOV) is also used. Finally, inflation reflects the price distortion's effects on an economy and may affect economic and financial activities. Hence, the annual rate of GDP deflator as inflation (INF) is also used to control the estimations. Since the data are in various scales, this study uses natural logarithm transformation for consistent results. The coefficients are considered as their elasticities.

Econometric Approaches and Empirical Models
The study uses a dynamic regression model. The bound test approach under the Autoregressive Distributed Lag (ARDL) model developed by (Pesaran and Shin 1998;Pesaran, Shin, and Smith 2001) is used to confirm cointegrating form and long-run level relationship whereas error correction model (ECM) is being used for short-run causality and confirmation of speed of adjustment towards long-run equilibrium.

Bound Test Approach Under ARDL Model
The ARDL model is a standard time series model that examines the relationship between a dependent variable and independent regressors both contemporaneously and historically by using In particular, this study's principle estimation model assumes that financial development is the function of economic growth in Nepal. Thus, real GDP per capita is the dependent variable, and the four proxies of financial development (used separately) and other control variables are explanatory variables.
Thus, the principal estimation model for economic growth under ARDL (p, q, r, s, t, u) is expressed by: On the contrary, the reverse estimation model assumes that economic growth is the function of financial development. Thus, the four proxies of financial development are used separately as the dependent variable, and real GDP per capita and other control variables are explanatory variables.
Once the cointegrating relationship between a dependent variable and independent regressors are confirmed, then equation 3 and 4 took the following form of the level relationship for the long-run.

Error Correction Model (ECM) Under the ARDL Model
If cointegration exists between two variables, there might exist at least one directional causality or bi-directional causality (Engle and Granger 1987). Therefore, this study further confirms the short-run causality between proxies of financial development and economic growth using the error correction model (ECM). The ECM is restricted to only two critical financial development variables, used separately and the proxy of economic growth. It provides partial information for adjustment and allows to estimate short-run elasticities. However, the estimation of short-run coefficients on error correction model (ECM) heavily depends on the optimum lags selected for the estimations.
Thus, The ARDL estimation equations of 3 and 4 took the following ECM equations for the short-run relationship.
In which, ∆ represents the first differenced value. The coefficients  1−  6 and  1−  6 provides the coefficients for short-term. p, q, r, s, t, u represents the number of lags selected automatically based on Akaike (1974) Information Criteria (AIC). ECT1 t−1 and ECT2 t−1 represent the one-period lagged value of error correction terms. Equation (7) gives the idea about the short-run causality of financial development and other control variables to economic growth and equation (8) gives the idea about the short-run causality of economic growth and other control variables to financial development. Finally, the coefficient of one period lagged value of ECT confirms the long-run causality and specify the speed of adjustment towards equilibrium.  The maximum domestic credit to the private sector is 81%, and broad money is 109% of GDP. The level of savings also varies across time from 4% to 16% of GDP. Trade openness varies from 21% to 45% of GDP. The government final consumption expenditure has remained between 8% to 12% of GDP. The maximum inflation exists at 18%, and the lowest is 3 % over the period.

Descriptive Statistics
It indicates that Nepal has gone through many volatile movements but not with the severe inflationary or deflationary condition. So, the result obtained from the estimated model using these variables gives a clear picture of Nepal's finance-growth relationship. Table 3 presents the correlation matrix. Each proxy of financial development is positively correlated and significantly correlates with real GDP per capita. So, they are used one by one separately. Other control variables, i.e. size of government and trade openness, are also positively correlated with real GDP per capita, but gross domestic savings and inflation negatively correlate with real GDP per capita. The correlation between the proxy of trade openness with proxies of financial development and economic growth may show a multicollinearity issue. So, the Lagrange Multiplier (LM) test is performed after the estimations to detect the multicollinearity issues.

Test of Stationarity
This study uses multivariate time series data under the ARDL estimations. ARDL model does not require selected variables to be the same order of integration. They could be either stationary or non-stationary at their level values. However, all the variables must have the same order of integration at the maximum of their first differenced values (Pesaran, Shin, and Smith 2001). This study confirms the stationarity and non-stationarity features of the series with the help of thee mostly and widely used unit root test methods. They are Augmented Dickey-Fuller (ADF) test, Phillips and Perron (PP) test, and Kwiatkowski, Phillips and Schmidt-Shin (KPSS) test. Table  model specification considers intercept only, and the second specification considers both trend and intercepts. Results show that the selected variables have mixed properties (stationary and nonstationary) in their level values. However, they all are stationary at their first differenced values in intercept specification. These results indicate that variables have the same order of integration at their first differenced value, i.e. I(1). So, the results provide sufficient backup to use ARDL bound test method for above-designed equations.  Table 5 reports the summary of results of the ARDL bound test approach for the cointegration. The results indicate that the calculated F-statistic is higher than the upper bound value when the proxy of economic growth, i.e. real GDP per capita is the dependent variable in four models of principle estimation equations. It indicates that financial development and other regressors have a long-run relationship with economic growth of Nepal. On the other hand, calculated F-statistic is higher than the upper bound value, when four proxies of financial development are used as the dependent variable separately in the three out of four models of the reverse estimation equations. It indicates that economic growth and other regressors have a longrun relationship with three proxies of Nepal's financial development. Therefore, the bound test approach of cointegration under the ARDL model confirms that Nepal's financial development and economic growth have two-way cointegration vector in the long-run. The level relationship further confirms their significance and direction of causality between them in the long-run.

Estimation Results
After confirming the cointegrating relationship between proxies of financial development and economic growth, the long-run level relationship is estimated. The real GDP per capita is the dependent variable in principle estimation models, and four proxies of financial development are the dependent variables in the reverse estimation models. The long-run results of principle estimation models are reported in table 6, and the results of reverse estimation model are reported in table 7, respectively. LnGDPPC LnGDS LnTRD LnGOV LnINF Notes: ***, ** and * indicates significance at 1%, 5%, and 10%, respectively. Source: Author's calculation.

Principle Estimation Model: Economic growth is the function of financial
development. Table 6 reports the long-run estimation results of equation (5), where the first differenced value of real GDP per capita (LnGDPPC) is the dependent variable as a measure of economic growth. The results indicate that all the proxies of financial development are positive and significant to cause real GDP per capita except broad money. It indicates that higher the level of financial development, higher would be the economic growth in the long-run. For example, one unit increase in domestic credit to private sector ratio to GDP causes 0.993-unit increase in real GDP per capita in the long run. Therefore, economic growth is financial development elastic in Nepal. Results also indicate that gross domestic saving is insignificant to cause real GDP per capita.
Trade openness is mostly negative and significant to cause real GDP per capita in model 1 and 2.
One possible reason behind this could be the import-based economic feature of Nepal. However, trade openness is insignificant to cause real GDP per capita in model 3 and 4. Size of the government is negative and significant to cause real GDP per capita in the long run in model 1 and 2. Inflation has a mostly negative but insignificant impact on real GDP per capita of Nepal. The short-run estimation results of equation (7) are reported in Table 7. The first differenced value of real GDP per capita (LnGDPPC) is the dependent variable, and four different proxies of financial development and other control variables are independent regressors. The estimation results indicate that none of the proxies of financial development causes real GDP per capita in short-run. It means that financial development is not short-run elastic to the economic growth of Nepal. The results also indicate that one year lagged value of trade openness has a positive and significant impact on real GDP per capita. It means trade openness has short-run positive effects on economic growth in Nepal. However, the size of government and inflation does not significantly affect economic growth in the short-run except model 3.

Reverse Estimation Model: Financial development is the function of economic
growth. Table 8 reports the long-run estimation results of equation (6) where the first differenced value of three proxies of financial development (LnDCP,LnDCB, LnBM) are the dependent variables. The results indicate that real GDP per capita is positive and significant to cause all proxies of financial development in Nepal. It indicates that higher the level of economic growth, higher would be financial development in the long-run. For example, one unit increase in real GDP per capita causes 0.868-unit increase in the domestic credit to the private sector ratio to GDP in the long run. Therefore, financial development is also economic growth elastic in Nepal. Results indicate that gross domestic saving is negative and significant to cause financial development in model 1 and 2. It indicates that savings are not mobilized through the financial system for credit expansion in Nepal. Trade openness is significant to cause proxies of financial development in model 1 and 2 but insignificant in model 3. The government's size is significant and positive to cause financial development proxies in model 1 and 2. Inflation is insignificant to cause all proxies of financial development. It indicates that Nepal has not faced hyperinflation by which the financial activates are not adversely affected in Nepal. Notes: ***, ** and * indicates significance at 1%, 5%, and 10%, respectively. Standard errors are in parentheses. All are transformed into a natural logarithm. Source: Author's calculation.
The short-run estimation results of equation (8) are reported in Table 9 where the first differenced value of four different proxies of financial development (LnDCP, LnDCB, and LnBM) are used as the dependent variable, separately, and real GDP per capita along with other control variables are independent regressors. The estimation results indicate that real GDP per capita is significant and positive to cause all financial development proxies, except broad money in the short-run. It indicates that if the economy grows, there could be quick demand for credit in shortrun in Nepal. However, gross domestic savings has adverse effects on financial development in the short-run as its first differenced value has adverse and significant effects on two proxies of financial development in model 1 and 2. The merchandise trade has a significant and positive effect on financial development in the short-run, but one period lagged value of merchandise trade has adverse effects on economic growth in model 1 and 2. The inflation also has a negative and significant impact on financial development in model 1. It means that financial development reacts negatively with inflation in the short-run. However, the size of government does not affect the financial development in the short-run.  Gregorio and Guidotti (1995); Jung (1986); and Kar et al. (2011). All measures of financial development are significant and positive to enhance economic growth in the long run.

Conclusion and Policy Recommendation
Thus, the study concludes that the Nepalese financial sector is well-functioning, especially with the backbone of financial reform and liberalization enacted after the 1980s.
Additionally, the positive and significant coefficients of real GDP per capita to cause financial development proxies in the reverse direction also indicate that economic growth is equally important to enhance Nepal's financial development. It also justifies that economic growth triggers financial development in the preliminary phase of economic development, as Hassan et al. (2011) concluded. The long-run coefficient of the domestic credit to the private sector and domestic credit to the private sector by banks ratios to GDP to cause real GDP per capita are very close. It justifies that the Nepalese financial sector is dominated by the banking sector rather than the non-banking financial sector. So, there is a large space to develop non-banking financial sectors in Nepal. The one-directional reverse causality from real GDP per capita to proxies of financial development in the short-run indicates that there could be quick demand for financial intermediation if the economic growth rises in the economy.
Trade openness is mostly negative and significant to cause real GDP per capita of Nepal.
The reason could be Nepal's import-based economic features, mostly triggered by the higher import of consumable goods with the backup of remittance inflows. Therefore, the trade integration policies must be guided by the export promotion and import substitutions economic policies in Nepal. However, trade openness is positive and significant to cause the proxies of financial development in Nepal. It indicates that trade liberalization might boost the financial sector in developing countries.
The gross domestic savings ratio to GDP, which is most volatile throughout the study period is insignificant to cause real GDP per capita in Nepal. It indicates that Nepal's economic growth does not depend on the level of domestic savings of Nepal. Things are clear that volatile savings are not intended to divert into investment. The government should formulate long-run policies which support to mobilize domestic savings into investment activities to achieve steady growth. A better investment environment is a prerequisite condition for saving mobilization on an economy.
The negative and significant coefficients of gross domestic savings towards two proxies of financial development also indicate that Nepal's financial market does not benefit from gross domestic savings. Things are clear that the government should set policies to stabilized savings that can be converted into the capital formation through financial intermediaries.
Size of the government is mostly negatively and significant to cause economic growth in Nepal. It suggests that a reduction in general expenditure in the government budget tends to raise economic growth. This argument may suggest that the government should have a sensitive budget allocation on general administrative expenditure versus capital expenditures such as infrastructure, education, and health. However, the size of government is significant and positive to foster financial development in Nepal. It means fiscal expansion gives opportunities to expand credit activities in Nepal. However, over expansion of fiscal speeding through general expenditure might introduce volatility in an economy.
Inflation is insignificant to cause economic growth and financial development in Nepal. It suggests that Nepal has not faced hyperinflation during the study period. However, high inflation might narrow down the financial and economic activities of an economy. So, the government should have control over inflation to have a stable financial system in the economy.

Future Research
For future research and for addressing possible weaknesses of this study, it is possible to increase the number of observations beyond 32 years. Further, the study may be broken down into different phases, such as before and after liberalization to get a different conclusion. This study has not used any dummy variables to represent Nepal's significant political, economic, and fiscal structural breaks. So, the use of an appropriate dummy may give more robust results. The study prefers bank-based financial indicators to measure financial development. It can be extended by using indicators from the capital market also. The other aspects of financial development, such as access, efficiency, and stability, are also equally important, thus can be used. The economic growth can be measured by other indicators also. Finally, this study uses a dynamic regression model.
Further research can be done by using other econometric methods by developing various model specifications.