The financial system of every country exists to ensure efficient resource allocation, mobilization of capital for investment, facilitate monetary policy, enable financial stability, and help payment system. The benefits of a well-functional financial system tend to contribute to the fortune of economic growth and development (Guptha and Prabhakar, 2018). A financial system comprises banks, capital  stock  exchange  markets,  and  insurance companies. In both developed and developing economies, the advancement of the financial system and efficient financial market has been identified as a contributing factor to economic growth (Xu et al., 2021). The efficiency and effectiveness of financial systems and their impacts on economic growth can be assessed using different institutions and parameters. However, academics and  policymakers  have mainly focused on banks’ role in the financial stability of economies and their contributions to economic growth, with little emphasis on stock exchange markets (Beck and Cull, 2013; Türsoy and Faisal, 2018). For instance, empirical evidence shows that domestic bank credit, non-performing loans, and bank capital to asset ratio positively affect the economic growth of twenty-eight (28) European Countries (EU) (Prochniak and Wasiak, 2017). The evidence from the developing countries is different. Bank credit negatively affects the economic growth of sixty (60) developing countries (Narayan and Narayan, 2013). The situation is not different with research on Sub-Saharan African (SSA) countries. Most African countries’ financial systems are developed, and their structure totally relies on the banking industry with a more significant influence on the countries’ economic growth at the total or partial neglect of the stock exchange market (Ouedraogo and Sawadogo, 2022). Notwithstanding the overemphasis literature on the impact of the banking industry as a pivot in the financial system on economic growth, there is strong evidence of the crucial role of an efficient stock exchange market in economic growth (Guptha and Prabhakar, 2018). It is not yet clear if this similarly holds for developing SSA countries. Consequently, this study evaluates Ghana’s financial system in the quest to understand whether or not it contributes to economic growth. According to studies, Ghana’s financial system is at a growing stage, inefficient, and coupled with challenges that stem from inappropriate dealings and inefficiencies by players in the various industries that make up the sector (Sackey and Nkrumah, 2012; Ofori-Abebrese et al., 2017; Amenu-Tekaa, 2022). The account put forward portrays that the inefficiencies led to the collapse and outright revocation of the license of several banks (Amenu-Tekaa, 2022). However, the Bank of Ghana (BoG) accounts that the country’s financial system is robust and resilient to the extent that players are posing high growth potential amidst higher profit due to the recent clean-up of the financial sector (BoG, 2019). Moreover, there is empirical evidence of the impact of the banking industry, one of the blocks in Ghana’s financial system, to economic growth (Sackey and Nkrumah, 2012; Ibrahim and Alagidede, 2020). The bank-based financial system is less sensitive to economic growth than the equity-based market (Ergungor, 2008; Liu et al., 2022). This varied fragmented discourse on the financial system calls for more empirical interrogation to establish the relationship between Ghana’s financial system and economic growth. In doing so, this paper seeks to assess if Ghana’s financial system based on efficiency (EF), stability (ST), and depth (DT) causes economic growth. The study would serve three thematic purposes. First, its findings would inform policymakers; again, practitioners in the financial sector would take a cue in structuring their policies and deals to enhance economic development. Finally, it would enrich the diverse literature on the discourse on the financial system and  economic  growth and policy recommendations.

Theoretical review

 

No single economic theory explains the relationship between the financial system and economic growth. However, this paper explores three theoretical frameworks that underpin the relationship between the stock market as a proxy for the financial system and economic growth. The three theories are the efficient market theory, asset pricing theory and Fisher's Hypothesis. First and foremost, the efficient market theory has been reconnoitred theoretically and empirically to conclude how security prices behave in an efficient market in different financial systems. The theory of efficient stock assumes that the security prices change fully to reflect the available information about individual stock, the entire stock market, or the economy (Fama, 1970). Although the term "how efficient is the stock market” has received many critiques as limitations to the efficient market theory, many scholars and researchers have tested, proven and added to the evolving of the effectiveness of the theory assumptions (Rayball, 1989; Adam and Tweneboah, 2008; Malkiel, 2003). The hypotheses of the efficient market theory have been used to assess stock market efficiency as either strong, semi-strong or weak form and the extent of such form contributions to the financial system (Ozdemir, 2008; Mubarok and Fadhli, 2020). The contribution of the stock market's efficiency has been linked to security prices and economic growth (Lee and Lee, 2009). The performance of the security prices, which reflects the efficiency of the market, is evident with the market's annual stock turnover as available information (Fama, 1970; Osei, 2002). Most empirical studies of the efficient market theory used the stock turnover to measure market efficiency information (Ozdemir, 2008; Shah et al., 2019; Bouteska and Regaieg, 2020). 

 

Moreover, the asset pricing theory, as propounded (Ross, 1976) and deployed in empirical studies on the stock market (Osei, 2002; Adu et al., 2013), asserts that volatility in the prices of listed stocks on an exchange is affected by systematic factors as well as unsystematic factors. The systematic factors, as proffered by the theory, refer to macroeconomic factors inherent in an economy which influences the mechanism that takes place in determining asset prices, whilst the unsystematic factors are those industry-specific variables that come with the particular industry a firm chooses to invest (Chen et al., 1986). These two distinct factors play an instrumental role in determining the price of listed stocks. As mentioned earlier, the effectiveness and accuracy of the factors reflect the efficiency and stableness of a country's  financial  market  (Borio and Drehmann, 2009). The effectiveness of the interaction between the asset pricing mechanisms would entice foreign firms to enlist their securities on a country's stock market (Martinez-Jaramillo et al., 2016). The higher the number of expatriate firms listed on the exchange, the higher the depth of the financial market of such country and consequentially higher economic growth with ceteris paribus. The asset price theory does not provide a framework that suggests that macroeconomic factors affect stock price volatility rather than the asset price; the fisher hypothesis proposed that expected nominal annual stock returns are a sum of the real returns and expected rate of inflation (Srivastava, 2010).

 

Lastly, Fisher's hypothesis, also known as Fisher's theory, is an economic framework brought by Irving Fisher (1930) which asserts that the real interest rate is distinct from monetary mechanisms, precisely the nominal interest rate and the expected rise in the general price of goods and services(Fisher, 1930). The framework describes the linkage between real and nominal interest rates and inflation. According to Fischer (1993), the difference between the real interest rate and the expected level of inflation sum up to the real interest rate. Consequentially a rise in the inflation rate equals a fall in the real interest rate unless a commensurable increment occurs in the nominal interest rate (Fischer, 1993). This volatility necessarily affects the stability of the financial market of every country, however, at different degrees depending on the level of financial advancement and economic growth of the country in question (Borio and Drehmann, 2009). The extent of the volatility of interest rates and inflation influences stock price volatility and thereby determines the financial system's stability. The underpinning assumptions of asset pricing theory and Fisher's hypothesis clearly suggest the interplay of stock prices and macroeconomic factors that determine the stability of a country's financial system and general economic climate. The theoretical framework of Fisher's hypothesis, the efficient market theory and the asset prices theory underpins a country's depth of financial system. Moreover, stock price volatility and stock market turnover are rooted in the assumptions of the efficient market theory and asset prices theory. This study adopts these theoretical frameworks to test the extent of Ghana's financial system's depth, efficiency and stability using the stock market as a proxy. 

 

Emperical review

 

The existing literature on the financial system and economic growth demonstrates dimensions where possible literature gaps that needs attenction in Ghana’s financial system. The financial system of a developing financial market is characterized by various factors contributing to market performance and economic growth. The factors identified by existing literature as influencing market efficiency are appropriate legal and regulatory framework, the efficiency of the securities exchange commission, a dynamic stock exchange market, and the extent of available information on a firm’s performance (Williams and Hussein, 2018). The stock market is described as efficient when the asset price fully displays all the available information (Eakins and Mishkin, 2012). The level of stock market efficiency has been identified as a key contributing factor to the stock market performance and its contribution to the economic climate of the country of operations (Bodeutsch and Franses, 2015; Jayakumar et al., 2018). For instance, weak stock market efficiency has been significantly linked to different degrees of economic growth (Aktan et al., 2019). The general empirical studies on the Ghana Stock Exchange (GSE) have mainly focused on the performance of the listed firms with few emphases on the efficiency of the market and the possible contribution to economic growth. A corporate financial performance review on GSE-listed banks indicates that the bank’s internal variables tend to increase profitability (Oduro et al., 2019) while the unlisted banks’ financial performance is inverse with the level of leverage (Gadzo and Asiamah, 2018). With regards to market efficiency, the GSE is not efficient due to high transactional cost, low firm participation, and slow market competition (Ayentimi et al., 2013). The GSE inefficiency is insensitive to firms’ returns on the market (Awiagah and Sup Brian Choi, 2018). It has also been established that the GSE is inefficient due to the information release of the annual earnings and the general market climate of the GSE is inconsistent with Efficient Market Hypothesis (EMH) theory (Osei, 2002). The inefficient market of GSE has several implications for domestic and international investors and economic growth (Magnus, 2008). Moreover, the competition in the stock market turns to improve efficiency, thereby impacting the country’s economic growth (Cojocaru et al., 2016).

 

The extent of GSE efficiency, performance, and contribution to economic growth is evident in the market regulations and the enforcement of those regulations (Tagoe, 2019). In Ghana, the level of economic growth and the extent of influence of sectors such as the financial system varies based on the financial market’s performance and development. There is evidence of a strong effect of Ghana’s capital market performance on economic growth (Acquah-Sam and Salami, 2014). The dynamic shocks in Ghana’s financial system development tend to have short-run and long-run effects on the country’s economic growth (Ibrahim and Alagidede, 2020). Besides, the government’s policy decisions, such as monetary policy, have a significant positive impact on Ghana’s stock market and influence economic growth (Antwi et al., 2021). A critical analysis of literature across continents and countries shows that the extent of the financial system’s impact on economic growth is determined by the choices of the variables used in the measurement of economic growth and the proxy of the financial system, thus, resulting in different findings (Adu et al., 2013; Alexiou et al., 2018). As an element that determines economic growth, the financial system has been measured as domestic credit (Mishra and Narayan, 2015; Prochniak and Wasiak, 2017), bank non-performing loans, and market capitalization (Prochniak and Wasiak, 2017; Antwi et al., 2021). In almost all these empirical studies, economic growth is measured by Gross Domestic Product (GDP).

The study adopted causal research as the design where the hypothesized individual variables would be regressed on economic growth as the dependent variable. The various variables included in the study are selected based on the indicators employed to assess the financial system. The depth of Ghana’s financial system is measured by international debt securities to GDP. The efficiency of the financial system is measured as the stock market turnover ratio. The use of multi-proxies to measure the financial system is motivated by the literature, and studies assert that defining a single appropriate variable for financial system development is a daunting task faced by empirical studies (Türsoy and Faisal, 2018; Koço?lu and Ci?hangi?r, 2021; Kuranchie-Pong and Forson, 2022). In addition, the measuring construct for financial system stability is stock price volatility (Antwi et al., 2021), while the dependent variable, economic growth, is measured by GDP per capita annual growth, as used by Barradas (2020). The multidimensional measure gives a comprehensive and broad assessment of the financial system development encompassing depth, efficiency, and volatility (Beck and Cull, 2013). Yearly data for Ghana’s financial system was assembled from 2008 to 2020, spanning a period of 13 years of time series observations. This period also represents the time frame within which data for all other variables employed in the study were collected. This period was considered for the study since most indicators’ data were unavailable before 2008. Moreover, 2008 was the year the global economy experienced a severe financial crisis where the financial systems of emerging economies, including Ghana, were worst hit (Blankenburg and Palma, 2009). Again, data for variables that predate 2008 were not in full. Since then, Ghana’s financial structure has witnessed several structural adjustments, including recapitalization and stock market listing requirements (BoG, 2019). The study employed the Granger causality framework as the estimation technique for its analysis (Granger, 1969). The adoption of the method is based on its usage in the empirical literature (Türsoy and Faisal, 2018; Koço?lu and Ci?hangi?r, 2021; Kuranchie-Pong and Forson, 2022). Again, the decision to employ granger causality is premised on the relationship the study intended to test. Moreover, since the data for the studied variables were time series, employing granger causality would serve the purpose of other conventional estimation models. Also, the approach assumes that DT, EF, and ST granger cause economic growth. This condition makes it suitable for the study since it assesses the effect of the regressors on the regressand (Granger, 1969). Since the data used is time series data, the stationarity of the variables needs to be tested; ignoring this would lead to spurious problems (Jayakumar et al., 2018). The Augmented Dickey-Fuller (ADF) tests for unit roots to ensure the variables are integrated in the same order (Harris, 1992).

 

Where Yt is a Random Walk and assumes the forms as follows:

 

 

 

The Philip-Person (PP) statistical test is a different approach to checking the unit root of a dataset. However, the rationale governing this does not differ from the ADF method. Notwithstanding, the PP test adds a non-parametric statistical approach to cater for serial correlation in the residual and does not add the lagged difference into the model. Below is the notation of the PP test.

Y_t is a random walk and takes the form as follows:

 

The results of various statistical tests and their corresponding discussions and interpretations are presented in this study. For any quantitative inquiry, the data’s description and summary are paramount. According to the descriptive statistics in Table 1, financial system stability had a maximum of 13.255, while that of economic growth was 11.315. The maximum efficiency of the financial system was 10.083, while the depth of the financial system recorded 15.214 as the maximum. The mean for economic growth was 3.880, efficiency was 5.675, and that of stability of the financial system was9.225. In addition, the mean for the depth of the financial system was 7.466. In total, the most divergent data from the center of the distribution was depth, with a standard deviation of 5.036, while the least deviated variable for the average of the distribution was stability which registered a standard deviation of 1.730. Efficiency is  highly  skewed to the left, notwithstanding the financial system’s stability skewed to the right. Jarque’s test showed that the dataset is normally distributed. The study begins by adjusting a Vector Error Correction Model (VAR) by identifying the Data Generating Process (DGP) to detect the characteristics of all variables. This VAR requires the study to estimate if the model would incorporate an aspect of intercept and time trend. Thereafter, the sampled indicators being its GDP_G, DT,  EF, and ST, would be subjected to unit tests, and only variables that integrate in the same order would be subjected to cointegration analysis. The variables were subject to the stationarity test through the ADF test and PP stationarity test; relying on the automatic level of precision selection by Akaike Information Criterion (AIC) and Schwarzman Information Criterion (SIC), using the PP approach to select the maximum lag length. Since the visual plot is usually the initial step in time series analysis, the study plotted the stationarity of the variables at level specification, as shown in Figure 2. The charts clearly show that the variables have no stationarity at a level, which signifies that their respective means and variance are not constant.

 

This finding agrees with previous studies that found unit roots for variables at level (Forson and Janrattanagul, 2014). For instance, the graphs for GDP_G and DT showed a fluctuating movement throughout the study period. Contrastingly, the graph for ST showed ups and downs, while that of EF demonstrated a decreasing trend with its peak during the 2008 financial crisis and economic downturn.

 

Unit Root

 

In statistical analysis, whenever a time series data exhibits the absence of a unit root, the series is integrated of order 0 or I (0). Nonetheless,  if  a  series  is integrated at I (1), I (2), or I (3), it means stationarity can only be achieved at the first difference, second difference, or third difference, respectively. After employing ADF and PP tests which test the null hypothesis that there is unit root and stationarity, the study ascertained mixed results for intercept and intercept with a time trend. Using the ADF for intercept and intercept with time trend, only efficiency was found without a unit root at a 5% level. However, as a confirmatory test, the PP test was employed to ascertain stationarity, but none of the variables was stationarity at a level. This PP test outcome led to testing at the first difference using both approaches. The ADF test found the GDP_G intercept without a unit root at the first difference at a 5% significance level. Again, the intercept and time trend of EF had a p-value of 0.001. Therefore, the null is rejected, concluding that EF does not have a unit root at first difference. PP test found the intercept and time trend of efficiency to be stationary at first difference. The remaining variables were either at the unit root or not stationary. The result means that at a level, only EF had its means and variance constant, which implies that the financial system was efficient at the level while the efficiency of the system was higher at the first difference (Tables 2 and 3).

 

Granger causality test 

 

The presence of a unit root problem due to the rejection of   the  alternative  hypothesis  means that  the  series  is non-stationary. This outcome makes it inappropriate and inefficient if the study applies the conventional granger causality test or investigates causality between the independent and dependent tests, making it necessary to select an alternative: Toda Yamamoto granger causality. The Toda and Yamamoto granger causality test was developed to serve as an alternative and more robust approach that tests causality among variables regardless of  whether  the variables are non-stationary or at the unit root. Therefore, the causality among the variables was tested using the Toda and Yamamoto granger causality, as shown in Table 4. Per the Toda Yamamoto test output, the null hypothesis that ST does not cause GDP_G cannot be rejected; thus, ST does not cause GDP_G. This output implies that the stock market’s stability does not lead to economic growth. It violates the study’s assumption that the more stable the financial system is, the higher the economic improvement. The result is in line with the findings of previous studies (Alexiou et al., 2018; Ehigiamusoe and Lean, 2018). However, the same contradicts the findings of previous studies (Bodeutsch and Franses, 2015; Min et al., 2018). The efficiency of the financial market was found to be inversely related to economic growth in Ghana. This outcome is based on the failure to reject the null hypothesis that efficiency does not cause economic growth. No long-run or short-run causality is running from efficiency to economic growth. This result means that the past efficient level of the stock market does not lead to economic growth in Ghana, which rejects the study’s initial hypothesis. This discovery conforms with the findings of Bhattacharya et al., (2019). However, the result disagrees with the findings of Min et al. (2018) and Antwi et al. (2021), who found the efficiency of a financial market to predict economic growth. Again, the depth of Ghana’s stock market does not cause economic growth, and the null hypothesis is accepted due to a p-value of 0.121. The result means that the level of depth of  the stock market activities is, the slower the economic growth; thus, there is no causality from depth to economic growth. This outcome confirms the conclusions of Alexiou et al. (2018) and Ehigiamusoe and Lean (2018) that the financial system decelerates economic growth; the same finding contradicts the assertions of Barradas (2020) and Jayakumar et al. (2018).