Financial Innovation and Economic Growth: Empirical Evidence from Nigeria
Ahmed Adesete^{1}, Adamu Mohammed Auwal^{2}, O. Dauda Risikat^{3}
Abstract: The study examines the impact of financial innovation on the Nigerian economy using both quarterly and monthly series ranging from 2010 to 2020. Most of the previous studies did not disaggregate the financial innovative channels in Nigeria using a valuebased approach and also fail to capture the full data information as they annualized both the quarterly financial innovation data and quarterly GDP time series. The study adopted the autoregressive distributed lag model (ARDL) and the polynomial distributed lag mixed data sampling (PDL MIDAS) model as it is effective when combining low frequency and highfrequency data. The outcome of the ARDL showed that mobile payment platform in Nigeria had a positive significant effect both in the long run and short run while internet web transactions and POS transactions have a negative insignificant effect on the Nigerian economy in the short run and long run. The PDL MIDAS equation result, also confirms mobile payment transactions is positively related to Nigeria’s economic growth while POS transactions were negatively related to the Nigerian economy. The results of the study suggest that financial innovation in the Nigerian financial system has a crucial role to play in the nation’s economy which validates the financegrowth theory. The study, therefore, recommends that there is need to strengthen policies towards improving the performance of the financial innovative channels in Nigeria and improve the financial security of existing payment platforms.
Keywords: Financial transactions; Online payment; Economic growth; Mixed data sampling; ARDL
JEL Classification: C22; C51; G20; O30; O41
1. Introduction
Financial innovation has continued to shape many business activities all over the world. It creates new businesses and provides better medium of exchange for goods and services. It also leads to the invention of new financial technologies which accelerate the productivity of capital, reduce transaction costs and hence stimulate economic growth. Financial innovation refers to the process of creating new financial or investment products, services or processes. According to Tahir et.al (2018) financial innovation refers to the introduction of new financial instruments in financial intuitions and markets through new technologies. It includes process, product and institutional innovation. In modern economy, the mode of payments plays a key role in the smooth functioning of economic activities. There is no doubt to the fact that financial innovations have created an efficient system of payments which provides instantaneous settlement of financial transactions and expedites the exchange of goods and services in a prompt, protected, and reliable manner. According to Qamruzzaman and Jianguo (2017), financial innovation promotes economic growth through financial inclusion and the facilitation of international exchange of goods and services.
Recently, Nigeria experiences financial innovations especially in the banking sector which move the country from being an analogue nation to a technology hub. Payments platforms such as Automated Teller Machines (ATMs), Point of Sales (P.O.S) Machines, electronic payment system, mobile banking have emerged from the technological development of the finance sector. These payments systems have resulted in reduction of transaction costs and created more employment opportunities for many youths in the country. It has also led to the emergence of new enterprises that provides a financial technology platform that allows individuals and businesses to make and receive payments from anywhere. There is an increase in the number of financial technology innovations enterprises that are driving banking services in Nigeria. This has made life and economic activities easier for Nigerians. However, despite the importance of these financial innovations, the process of making payments online in Nigeria have been marred with many challenges.
Enormous empirical studies carried out in Nigeria shows that financial innovations via most electronic channels was found to contribute positively to the economic growth while some others conclude that it does not have a desired effect on the economic growth of Nigeria. Ozurumba and Charles (2019) examined the effect of financial innovation on the economic growth of Nigeria and concluded that the volume of financial transactions via NIBSS and Agent Banking platforms have positive but insignificant effect on economic growth of Nigeria, while the volume of transactions carried out through the Automated Teller Machine (ATM) have negative effect on economic growth. Meanwhile, John (2019) adopted Autoregressive Distributed Lag (ARDL) technique to explore relationship between the adoption of electronic payment system and the growth of the Nigerian economy and concluded that automated teller machine, mobile payment and webbased transactions have negative impact on the real GDP growth of Nigeria while POS transaction have positive impact on economic growth. Ogedebe and Jacob (2012) attributed the problems associated with financial innovation channels in Nigeria to unreliable network connectivity and cyber insecurity. BCBS (2017), Ahmed and Monir (2019) also identified challenges such as data security, high operational risk, vendor management risk, money laundering, liquidity risk hampering the effectiveness of financial innovation channels. These challenges had led many people to abandon the use of financial technology and innovation platforms offered by vendors and banks, preferring to withdraw and deposits money in the banking hall. Therefore, there is need to investigate the impact of financial innovation on the Nigerian economy to contribute to existing literatures in this line of knowledge.
This study therefore aims to fill the knowledge gap in existing literature in the following ways. Firstly, this paper employs a valuebased approach (as it captures both the local currency value and volume of financial transactions through financial innovation channels) by disaggregating financial innovation into five important financial innovation channels which are values of POS, internet web, cheques, mobile payments and ATM transactions to determine the relationship between financial innovation and the Nigerian economy. Secondly, the study employs an high frequency data (monthly and quarterly) to capture the objective of this paper as against annualizing the time series as used by most studies in this line of knowledge. Thirdly, this paper adopts a dynamic approach to explore the relationship between financial innovation and economic growth by employing a dynamic mixed data sampling (MIDAS) and the ARDL model. Lastly, the study uses the MIDAS model to capture the full time and data information of both the GDP variable (which is a quarterly series from first quarter 2010 to first quarter 2020) and financial innovation indicators (which is a monthly series from January 2010 to April 2020).
It is against this background that this study adds to the body of knowledge. This paper will help the policy makers to understand the role of financial innovation on the economy. Also, the empirical result of this paper will help identify the financial innovation channels that are important and that needs to be improved in order to aid the development of the financial sector and economic growth in Nigeria. In addition, this paper will help policy makers understand the trend of financial innovation in Nigeria so far with a view to identifying challenges facing by financial innovation system users in Nigeria and proffering possible solutions to the problems identified.
The broad objective of this paper is to examine the impact of financial innovations on the Nigerian economy. The primary objectives of this paper are to examine the impact of financial innovation through five indicators such as value of internet web transactions (VAIW), value of mobile payment transactions (VAMP), value of ATM transactions (VATM), value of POS transactions (VAPOS) and value of cheque transactions (VACT) on the Nigerian economy as well as to examine the trend of these financial innovation indicators in Nigeria.
In order to achieve this objective, the paper is structured as follows: Section 1 is the introduction, while section 2 provides an overview of the related literature reviews on the financial innovations as well as its impact on an emerging economy. In section 3 and 4, the methodology and results are respectively presented. The conclusion and recommendations from the study are presented in section 5.
2. Literature Review
Although evidence from literature have shown that financial innovation has played a vital role on the growth of any economy whether developed or developing. Moreso, there are quite a number of empirical studies on the impact of financial innovation on economic growth. However, the results obtained from many of these studies on various financial innovation variables have not being conclusive.
Okafor et al. (2017) emphasized the effects of online payment channels on the growth of the Nigerian economy. The study employed the Least Square (GaussNewton/Marquardt steps) based on vector autoregressive (VAR) system and Johansen cointegration test to explain the relationship among the variables. The study made use of data covering the period from 2009 to 2014 sourced from Central Bank database on financial innovation. The study used variables such as total volume of transactions via internet banking, volume of transactions through automated teller machine (ATM), and the total volume of Point of Sale (POS) transactions, while gross domestic product is proxied for economic growth. The results obtained from the Johansen cointegration test revealed the existence of long run relationship between economic growth and financial technology innovation in Nigeria, that is, both variables move together over a long period of time. The regression results revealed that financial technological innovation via automated teller machine (ATM) and internet banking have significant positive impact on economic growth in Nigeria, while financial technological innovation through point of sale (POS) payment channels have negative impact on economic growth in Nigeria. In other words, financial transactions or payments through automated teller machine (ATM) and Internet banking platforms influence economic growth in Nigeria while financial transactions or payment through the point of sale (POS) machine worsened the economic growth prospects of the country.
Qamruzzaman and Jianguo (2017) applied Autoregressive Distributed Lag (ARDL) bound test and the GrangerCausality based Error Correction Model (ECM) to investigate the existence of long run relationship among the variables used and also to find the direction of causality among the variables. The study used real GDP per capita as a measure of economic growth which is widely used by economic researchers. On the part of financial innovation, the study used two indicators as a measure, the first is the ratio of domestic credit to private sector and the second is the ratio of BroadtoNarrow Money (M2/M1). The results from the cointegration test revealed the existence of long run relationship between financial innovation and economic growth in Bangladesh. The results further revealed that financial innovation relating to both the ratio of domestic credit to private sector as well as the ratio of BroadtoNarrow Money have positive and significant effect on economic growth of Bangladesh. This implies that the Performance of the banking sector through financial innovation have significant influence on the economic growth in Bangladesh. The causality test revealed that there is evidence of causality from financial innovation to economic growth. This implies that any shock in financial innovation will bring about a positive impact on economic growth of Bangladesh.
Ravikumar et.al (2019) assess the impact of digital payments on the growth of Indian economy. The study used the Ordinary Least Square (OLS), Autoregressive Distributive Lag (ARDL) cointegration approach and ARDL bound test techniques for the analysis. Digital payments were measured using Real Time Gross Settlement (RTGS), Clearing Corporation of India Ltd (CCIL) operated system, paper clearing, retail electronic clearing, card payments, and Prepaid Payment Instruments (PPIs). While real GDP is used as a measure for economic growth during the periods of 2011 to 2019. The OLS results show that only retail electronic payments as a measure of digital payments influences the real GDP significantly. All other variables used as measures of digital payments do not have significant impact on the real GDP in India. Moreover, the results obtained from the ARDL bound test revealed that digital payments do not have significant impact on the economic growth of India in the long run.
Ozurumba and Charles (2019) employed Ordinary Least Square (OLS) technique to examine the effect of financial innovation on the economic growth of Nigeria. The study used data on Nigeria Interbank Settlement System (NIBSS) transactions, Amount of Automated Teller Machine (ATM) transactions and the transactions of the Agent Banking function over the period 2012 to 2018. The results from the regression analysis revealed that the volume of financial transactions via NIBSS and Agent Banking platforms have positive but insignificant effect on economic growth of Nigeria, while the volume of transactions carried out through the Automated Teller Machine (ATM) have negative effect on economic growth and statistically significant at 5% level.
John (2019) adopted Autoregressive Distributed Lag (ARDL) technique to estimate the relationship between the adoption of electronic payment system and the growth of the Nigerian economy. The study used data on the volume of transactions carried out on webbased transactions (WBT); point of sales (POS); automated teller machine (ATM); Mobile payment (MOP); interbank money transfer transactions (INTERBANK) covering the period 20122017. The results from the empirical findings revealed that automated teller machine, mobile payment and webbased transactions have negative impact on the real GDP growth of Nigeria. This means that ATM, MOP, WBT transactions contribute negatively to economic growth in Nigeria. The results further revealed that POS transaction has positive impact on economic growth, while interbank transaction has an insignificant impact on the growth of Nigeria.
Moreover, Chukwunulu (2019) investigated the effect of financial innovation on the growth of the Nigerian economy. The study used data on GDP growth and the four major emoney transactions used in Nigeria as proxies for financial innovation, namely: amount of automated teller machine (ATM) transactions, amount of web transactions, amount of pointofsale (POS) transactions, and the amount of mobile payments transactions for the period of 20082017. The research employed the Generalized Methods of Moments (GMM) since the time frame is short. The results obtained from the data analysed show that all the four major emoney transactions used in Nigeria (i.e amount of ATM transactions, amount of internet transactions, amount of POS transactions, and the amount of mobile payments transactions) have significant positive effect on the GDP growth in Nigeria. This indicates that financial innovation has impacted positively on the economic growh of Nigeria over the period 20082017 which contradicts the results obtained by (John, 2019) which show that automated teller machine, mobile payment and webbased transactions have negative impact on the real GDP growth of Nigeria.
From the empirical studies reviewed, most of the findings showed that various channels of online financial transactions give mixed results in different countries. For instance, some studies carried out in Nigeria revealed that financial innovations via most electronic channels were found to contribute positively to the economic growth while some other studies conclude that it does not have the desired effect on the economic growth of the Nigerian economy. The results from this study will serve as an empirical evidence showing the impact of financial innovation on Nigeria economy and will be useful to policymakers in formulating policies for the realisation of allimportant inclusive growth necessary for sustainable development
3. Research Methodology
3.1. Theoretical Framework
This paper was anchored on the financegrowth theory, theory of Task Technology Fit (TTF) which was developed by Goodhue and Thompson in 1995 and the Romer endogenous growth theory.
The Task Technology Fit (TTF) theory states that information technology mostly has a positive relationship with individual performance especially if the task to be performed with the technology matches the individual’s skills. This indirectly indicates that the efficiency of financial innovation depends on the capability of the user. This theory is relevant to this paper because diverse range of financial innovation systems such as POS, internet web and mobile payment channel and its effectiveness on financial sector development and economic development generally. In short, applying the TTF concept to financial innovation concludes that for financial innovative systems such as the epayment channels to have significant impact on individuals and the aggregate economy, the users must be capable of using the system and it must match their need.
Secondly, the financegrowth theory posits that financial development contributes to productivity through ‘demand following’ and ‘supply leading’ effect. This theory can be traced back to Bagehot work in the 1870s. Accessibility and affordability of funds is a precondition for boosting economic growth under this theory. Bagehot emphasized the role of the financial institutions and systems in pooling and allocating financial resources to the most profitable enterprises or sectors. Efficient allocation of this financial resources will have a multiplier and positive spillover effect on economic growth. On the demand side, aggregate demand will increase due to accessibility to more finance which is expected to increase investment expenditure. From the supply side, boost in accessibility to finance by various sectors and investors will enhance output of firms and different sectors of the economy. This is expected to positively contribute to employment, human welfare, firm’s profitability and overall aggregate demand.
Finally, from a macroeconomicsbased approach, the relationship between finance and economic growth can be explained with the aid of Lucas and Romer endogenous growth theory.
Assume production function,
Equation one is an endogenous production function which shows that output is determined by capital and augmented labour. K is assumed to represent financial capital while Y is proxy to real gross domestic product (RGDP).
 Technology constant
NPopulation which is assumed to be exogenous such that output is a function of only financial capital.
Rearranging (1) to separate the constants
Divide both sides of equation (3) by A,
Output manufactured (growth) with the aid of technology
Equation (6) represents the transformed output function of equation (1) which now indicates that technological augmented output is a function of constant population growth (N) and technological augmented financial capital (AK).
Financial capital innovation
Equation 8 shows that output (RGDP) is determined by population which is assumed to be exogenous to the model and technological augmented financial capital (financial capital innovation.
Taking the natural logarithm of the LHS (left hand side) and RHS (right hand side) of equation (8) to linearise the equation.
Population is exogenous:
Let,
Equation (11) is the final equation that shows the relationship between output (y) and financial innovation (k). According to the equation, there is a positive relationship between financial innovation and economic growth.
Conclusively, the TTF theory posits that financial innovative system will only have an impact on individual performance and the aggregate economy only if users of these financial innovative channels are capable of using it thereby matching capability to effectiveness. Also, the finance growth theory proves to explain the role of finance on economic growth through the development of the financial system of a particular country. This financegrowth theory shows that financial development has a very important role to play in ensuring easy accessibility and availability of fund to investors which will in turn increase investment expenditure thereby boosting aggregate demand and economic growth. Lastly, from the macroeconomic growth model, the assumed output function under the assumption of exogenous population growth also concludes that there is a positive relationship between financial capital innovation and economic growth. In short, this theory helps explain the role of financial innovation such as the POS, mobile payments, internet web transactions, mobile payments on an economy through different approach but the same conclusion.
3.2. Model Specification
The model used for this paper was developed on the assumption from the theoretical framework that financial innovation influences economic growth. This model is in line with Chukwunulu (2019) who specified that Nigeria’s GDP growth is determined by financial innovation variables such as value of ATM transactions, web transactions, POS transactions and mobile payments. This paper model is however specified thus merging this assumption and Chukwunulu (2019) model:
RGDP = f (Financial innovation)
Financial innovation is proxy by variables such as the value of ATM transactions (VATM), cheque transactions (VACT), internet web transactions (VAIW), POS transactions (VAPOS) and mobile payments (VAMP)
RGDP = f (VATM, VACT, VAIW, VAPOS, VAMP)
Note: all variables were estimated in their logged forms.
The ARDL model estimated for this study is specified as follows:
The full ARDL equation specification for this paper is as follows:
The MIDAS equation is specified thus:
Applying the generalized MIDAS model in (15) to the model specified for this paper,
Recall:
k = financial capital
y = real gross domestic product (RGDP)
Let k proxy financial innovation
There are two approaches to estimating mixed frequency model (equation 19) which are the individual coefficient and aggregation approach. Also, the mixed data sampling (MIDAS) model offers several weighing functions which helps to reduce the number of parameters in the model by placing restrictions on the effect of high frequency on the low frequency variables at various lag periods. Equation 19 can be expressed explicitly to show the high frequency variable in relation to the low frequency variable. Financial innovation variable such as VACT, VAMP, VAIW, VAMP and VAPOS are the high frequency variables (monthly time series) while the real gross domestic product (RGDP) is the low frequency variable (quarterly time series).
Monthly series on financial innovation variables with rth frequency lags
It is important to note that the Almon polynomial distributed lag weighting function was adopted in this paper to estimate the mixed data sampling (MIDAS) model for the following reasons:
It helps reduces the effect of multicollinearity. (Judge et al. 1980)
It reduces the number of parameters to be estimated when using a dynamic linear model (Maddala, 1977).
Applying the Almon PDL weighting function to equation (20):
Z = weight
Equation 21 represents the final MIDAS model after applying the Almon polynomial distributed lag (PDL) weighing function to estimate equation 19 and equation 21 is the Almon weight applied to equation 20 to derive equation 22.
Generally, a MIDAS specification captures a very rich dynamic property of the highfrequency process in a very simple and parsimonious fashion. In this paper we will be using monthly data of financial innovation proxies such as value of POS transactions (VAPOS), value of cheque transactions (VACT), value of internet web payment transactions (VAIW), value of mobile payments transactions (VAMP) and value of ATM transactions (VATM). But, monthly data on real gross domestic product (RGDP) is not available, hence, we will use a quarterly data on RGDP. Using a high frequency data for financial innovation and low frequency data for RGDP justifies the use of mixed data sampling (MIDAS) model to examine the relationship between financial innovation and the Nigeria’s economy.
3.3. Estimation Techniques
To ensure stationarity of the time series data used, Augmented Dickey Fuller and the KPSS unit root test was carried out for each of the variables both at level and first difference. Cointegration test using the ARDL bound test was carried out to verify the existence of longrun relationship between the dependent variable and explanatory variables. The parameters of the equation were then estimated using both the ARDL and polynomial distributed lag mixed data sampling (PDL MIDAS) model. The MIDAS model was used because the series on value of POS transactions (VAPOS), value of cheque transactions (VACT), value of internet web payment transactions (VAIW), value of mobile payments transactions (VAMP) and value of ATM transactions (VATM) which were used as proxies to financial innovation was available in monthly time period while the data on real gross domestic product (RGDP) is in quarterly time period from its source. Meanwhile, the financial innovation variables were also disaggregated into quarterly series by taking the sum of per threemonth (say from January to March, April to June etc) data which makes up to a quarter. This disaggregated data was used alongside the quarterly series on RGDP to estimate the parameters of the ARDL model. Finally, post estimation tests such as normal distribution, serial correlation, heterosedasticity, stability and misspecification bias tests were performed on the residuals of the equations to ascertain if they conform to the normal assumptions.
3.4. Sources and Measurement of Data
Variable Name 
Description 
Source 
Type of Variable 
Apriori Expectation 
RGDP 
Real gross domestic product in billion Naira used as a proxy to performance of the Nigeria economy ranging from Q1’2010 to Q1’2020. 
CBN Database 
Dependent (Quarterly) 

VAPOS 
Value of POS transactions in billion Naira proxy to financial innovation ranging from January 2010 to April 2020. 
CBN Database 
Independent (Monthly) 
Positive 
VACT 
Value of cheque transactions in billion Naira proxy to financial innovation ranging from January 2010 to April 2020. 
CBN Database 
Independent (Monthly) 
Positive 
VATM 
Value of ATM transactions in billion Naira proxy to financial innovation ranging from January 2010 to April 2020. 
CBN Database 
Independent (Monthly) 
Positive 
VAIW 
Value of internet web transactions in billion Naira proxy to financial innovation ranging from January 2010 to December 2019. 
CBN Database 
Independent (Monthly) 
Positive 
VAMP 
Value of mobile payment transactions in billion Naira proxy to financial innovation ranging from January 2010 to December 2019. 
CBN Database 
Independent (Monthly) 
Positive 
4. Estimation and Interpretation of Results
4.1. PreEstimation Analysis
4.1.1. Trend Analysis
Source: Drawn by the author with the aid of WPS Spreadsheet 2020.
Figure 1 and 2 show the monthly and quarterly trend of the series for Value of Cheque Transactions (VACT), Value of ATM Transactions (VATM), Value of POS Transactions (VAPOS), Value of Internet (Web) Transactions (VAIW) and Value of Mobile Payment Transactions (VAMP) as well as the quarterly trend of real GDP of Nigeria.
From the monthly trend, the value of cheque transactions was relatively high between January 2010 and December 2011 recording an all time high of N2129.01 billion in November 2011. Meanwhile, the value of cheque transactions fell sharply by 355.6% to N449.13 billion in January 2012 from N2046.17 billion in the previous month. Similarly, value of POS transactions, ATM transactions and Mobile payment transactions fell sharply to N400 million, N127.59 billion and N240 million respectively in January 2012. Values of internet web transactions, POS transactions, ATM transactions and Mobile payment transactions has been rising and stagnating since 2012 but value of cheque transactions has been declining since 2012, with average VACT falling to N373.47 billion in 2019 from N621.8 billion in 2012. So far, between January and March in 2020, value of POS and ATM transactions advanced 17.68% and 1.96% to N368.86 billion and N516.34 billion respectively. This is owing to increase in the use of mobile payment platforms and ATM during the governmentimposed lockdown in response to the COVID19 outbreak in Nigeria. Meanwhile, there was a sharp drop in the value of cheque transactions to N10.3 billion in April 2020 from N361.71 billion in March. In the same vein, value of POS and ATM transactions declined in April, 2020 from the previous month.
4.1.2. Descriptive Statistics
Table 1. Descriptive Statistics for Quarterly Data

RGDP 
VAPOS 
VATM 
VACT 
VAIW 
VAMP 
Mean 
16247474 
238.4995 
984.4802 
2253.552 
41.77375 
226.237 
Median 
16218542 
96.34 
1011.48 
1654.32 
22.575 
96.03 
Maximum 
19750935 
1008.32 
1832.55 
6109.31 
221.52 
1687.1 
Minimum 
12583478 
1.87 
62.59 
1087.96 
3.37 
0.87 
Std. Dev. 
1724008 
296.3122 
536.9772 
1530.918 
52.35771 
379.0816 
Skewness 
0.138808 
1.258923 
0.101192 
1.495748 
2.172447 
2.647751 
Kurtosis 
2.414792 
3.343321 
1.738025 
3.54331 
7.167124 
9.4867 
JarqueBera 
0.716713 
11.03143 
2.790631 
15.79223 
60.40506 
116.866 
Probability 
0.698824 
0.004023 
0.247755 
0.000372 
0.0000 
0.0000 
Observations 
41 
41 
41 
41 
40 
40 
Source: Author’s Computation with the aid of Eviews 10
Quarteronquarter, the average real GDP for Nigeria between Q1’2010 and Q1’2020 is N16.25 trillion. The lowest and the highest real GDP up till Q1’2020 is N12.58 trillion (in Q1’2010) and N19.75 trillion (in Q4’2019) respectively. The skewness statistics reveals that all the series are skewed to the right (positive skewness) except RGDP and VATM which is skewed to the left (negative skewness). Kurtosis also shows that all the quarterly series are leptokurtic (kurtosis value greater than 3) except RGDP and VATM which is platykurtic (kurtosis statistic less than 3). The Jarque bera test statistic indicates that RGDP and VATM series are normally distributed because the probability value is greater than 5%.
Table 2. Descriptive Statistics for Monthly Data

VACT 
VAIW 
VAMP 
VAPOS 
VATM 
Mean 
745.209 
13.92458 
75.41233 
81.05266 
327.9258 
Median 
537.7700 
7.5100 
31.6450 
33.7750 
320.8100 
Maximum 
2129.010 
168.200 
743.170 
372.69 
670.980 
Minimum 
10.3000 
0.8800 
0.2200 
0.0400 
18.4800 
Std. Dev. 
515.0034 
23.07263 
127.3332 
99.54974 
178.4469 
Skewness 
1.532251 
5.105949 
2.861896 
1.252399 
0.054845 
Kurtosis 
3.841142 
33.59291 
11.60668 
3.370678 
1.811809 
JarqueBera 
52.17656 
5201.046 
534.1832 
33.12562 
7.356451 
Probability 
0.00000 
0.00000 
0.00000 
0.00000 
0.025268 
Observations 
124 
120 
120 
124 
124 
Source: Author’s Computation with the aid of Eviews 10
On the other hand, table 2 shows the descriptive statistics for the monthly series of the financial innovation indicators such as Value of Cheque Transactions (VACT), Value of ATM Transactions (VATM), Value of POS Transactions (VAPOS), Value of Internet (Web) Transactions (VAIW) and Value of Mobile Payment Transactions (VAMP). Average value for VACT, VAPOS and VATM between January 2010 and April 2020 is N745.21 billion, N81.05 billion and N327.93 billion respectively while the average for VAIW and VAMP between January 2010 and December 2019 is N13.92 billion and N75.41 billion respectively.
Table 3. Maximum and Minimum Values of Financial Innovation Indicators with their Corresponding Dates

Maximum 
Date 
Minimum 
Date 
VACT 
2129.01 
November, 2011 
10.3 
April, 2020 
VAIW 
168.2 
November, 2018 
0.88 
January, 2010 
VAMP 
743.17 
December, 2012 
0.22 
January, 2010 
VAPOS 
372.69 
December, 2012 
0.04 
January, 2012 
VATM 
670.98 
November, 2017 
18.48 
February, 2010 
Source: Author’s Computation with the aid of Eviews 10
4.1.3. Unit Root Test
H_{0}  The series are stationary
H_{1}  The series are not stationary
Table 4. The ADF and KPSS Unit Root Test Result
Variables 
Augmented Dickey (ADF) Fuller Test Result 
KwiatkowskiPhillipsSchmidtShin (KPSS) Test 

Level 
First Difference 
Level 
First Difference 

RGDP 
0.2600 
0.4224 
0.0000 (0.8407 > 0.4630) 
0.5755* (0.3246 < 0.4630) 
VAPOS 
0.6762 
0.0000* 
0.0000 (1.3029 > 0.4630) 
0.2833* (0.2455 < 0.4630) 
VATM 
0.6297 
0.0000* 
0.0000 (1.1415 > 0.4630) 
0.0817* (0.9169 > 0.4630) 
VACT 
0.9485 
0.0000* 
0.0000 (1.1361 > 0.4630) 
0.2277* (0.2583 < 0.4630) 
VAIW 
0.6594 
0.0000* 
0.0000 (1.2362 > 0.4630) 
0.4447* (0.0777 < 0.4630) 
VAMP 
0.6297 
0.0000* 
0.0000 (1.2611 > 0.4630) 
0.0510* (0.2106 < 0.4630) 
Source: Author’s Computation with the aid of Eviews 10
* denotes acceptance of stationarity at 5% significance level
Values in parenthesis indicates LM stat and critical values at 5% significance level respectively
Table 4 shows the ADF and KPSS unit root test result for real GDP, value of cheque transactions (VACT), value of ATM transactions (VATM), value of POS transactions (VAPOS), value of internet (Web) transactions (VAIW) and value of mobile payment transactions (VAMP).
According to the ADF unit root test result, all series were only stationary at first difference except the series for real GDP which was shown by ADF not to be stationary at level (I(0)) and first difference (I(1)). This is because the probability value for VACT, VATM, VAPOS, VAIW and VAMP was less than 5% at first difference thereby accepting the null hypothesis that the series were stationary.
On the other hand, the KPSS test result indicates that all the series used in this study such as RGDP, VACT, VATM, VAPOS, VAIW and VAMP were stationary only at first difference (I(1)) because the KPSS probability values were greater than 5%. According to the KPSS test, the null hypothesis is not rejected for all the series at first difference only thereby concluding that all series in this paper are stationary at first difference only.
4.1.4. Cointegration Test
H_{0}  There is no cointegration
H_{1}  There is presence of cointegration
Table 6. ARDL Bound Test Result
Test Statistic 

Fstatistic 
Value 
k 
5.447418 
5 

Critical Value Bounds 

Significance 
I0 Bound 
I1 Bound 
10% 
2.26 
3.35 
5% 
2.62 
3.79 
2.50% 
2.96 
4.18 
1% 
3.41 
4.68 
Source: Author’s Computation with the aid of Eviews 10
Table 6 shows the ARDL bound test result of the ARDL model. This result indicates that there is presence of cointegration (that is long run relationship) between the dependent variable (RGDP) and financial innovation (VACT, VAMP, VAIW, VATM and VAPOS). Variables are cointegrated because the F statistic value of the ARDL bound test is greater than the upper bound value of the critical table bounds value at the significance levels indicated in the table.
Table 7. Johansen’s Trace Test Result
Unrestricted Cointegration Rank Test (Trace) 

Hypothesized No. of CE(s) 
Eigenvalue 
Trace Statistic

0.05 Critical Value 
Prob.** 
None * 
0.835446 
186.5947 
95.75366 
0.0000 
At most 1 * 
0.729291 
118.0231 
69.81889 
0.0000 
At most 2 * 
0.546722 
68.36799 
47.85613 
0.0002 
At most 3 * 
0.444001 
38.30048 
29.79707 
0.0042 
At most 4 * 
0.279959 
15.99489 
15.49471 
0.0420 
At most 5 
0.088325 
3.513911 
3.841466 
0.0609 
Source: Author’s Computation with the aid of Eviews 10
Trace test indicates 5 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
Table 8. Johansen’s Maximum Eigenvalue Test Result
Unrestricted Cointegration Rank Test (Maximum Eigenvalue) 

Hypothesized No. of CE(s) 
Eigenvalue 
MaxEigen Statistic 
0.05 Critical Value 
Prob.** 
None * 
0.835446 
68.57161 
40.07757 
0.0000 
At most 1 * 
0.729291 
49.65506 
33.87687 
0.0003 
At most 2 * 
0.546722 
30.06751 
27.58434 
0.0235 
At most 3 * 
0.444001 
22.30559 
21.13162 
0.0341 
At most 4 
0.279959 
12.48098 
14.2646 
0.0939 
At most 5 
0.088325 
3.513911 
3.841466 
0.0609 
Source: Author’s Computation with the aid of Eviews 10
Maxeigenvalue test indicates 4 cointegrating eqn(s) at the 0.05 level
* denotes rejection of the hypothesis at the 0.05 level
Table 7 and 8 shows the Trace and Maximum Eigenvalue test statistic of the Johansen cointegration test result. The cointegration test checks for the presence of long run relationship among the variables used in this study. Since all the variables used in this study were only stationary at I(1), it is necessary to test if the variables converge in the long run. The null hypothesis is rejected if there is presence of at least one cointegrating equation in the trace and maximum eigenvalue test statistic result, otherwise the null hypothesis is accepted.
Trace test shows the presence of five (5) cointegrating equations while the maximum eigenvalue indicates the presence of four (4) cointegrating equations. We however reject the null hypothesis and conclude that there is long run relationship between the variables used in this paper.
4.1.5. Lag Selection for ARDL Model
Table 9. VAR Lag Selection Criteria
Source: Author’s Computation with the aid of Eviews 10
The maximum lag length for the variables employed in this paper is one (1) according to the Schwarz criterion while other criterion indicates three (3) maximum lag length for the ARDL model estimated in this paper.
4.1.6. ARDL Model Selection
Figure 2. ARDL Model Selection Using the Akaike Information Criteria
Source: Author’s Computation with the aid of Eviews 10
From the VAR lag selection criteria, the maximum lag length for the variables is 3 as concluded by majority of the information criterion including the Akaike information criterion (AIC). The RGDP ARDL model was estimated with the aid of this lag lengths and the model with the highest AIC value was selected. From about 243 models that was estimated using combinations of different lags, the ARDL model (1, 0, 0, 0, 0, 0) was selected because it has the highest AIC values of all the models estimated.
4.2. Estimation and Interpretation of Results
Table 10. The RGDP ARDL Model (1, 0, 0, 0, 0, 0)
Variable 
Coefficient 
Std. Error 
tStatistic 
Prob. 
SHORTRUN EQUATION 

DLOG(VACT) 
0.010437 
0.038461 
0.271363 
0.7879 
DLOG(VAIW) 
0.003676 
0.021084 
0.174371 
0.8627 
DLOG(VAMP) 
0.054141 
0.02537 
2.134084 
0.0406* 
DLOG(VAPOS) 
0.019379 
0.03513 
0.551641 
0.5850 
DLOG(VATM) 
0.013566 
0.036687 
0.369775 
0.7140 
ECT(1) 
0.869761 
0.163778 
5.310611 
0.0000* 
LONGRUN EQUATION 

LOG(VACT) 
0.012 
0.044561 
0.269283 
0.7894 
LOG(VAIW) 
0.004227 
0.024325 
0.17377 
0.8631 
LOG(VAMP) 
0.062248 
0.032337 
1.924946 
0.0632** 
LOG(VAPOS) 
0.022281 
0.041019 
0.543196 
0.5908 
LOG(VATM) 
0.015597 
0.041371 
0.377007 
0.7087 
C 
16.265108 
0.486269 
33.448811 
0.0000* 
Rsquared 0.733842 Fstatistic 14.7049 Adjusted Rsquared 0.683938 Prob(Fstatistic) 0.0000 DurbinWatson stat 1.92122 
Source: Author’s Computation with the aid of Eviews 10
* significant at 5% and 10%
** significant at 10% only
DEPENDENT VARIABLE: DLOG(RGDP)
Table 10 shows the ARDL model comprising of the long run and short run equation result with logged form of real gross domestic product (RGDP) as the dependent variable and financial innovation proxy by VACT, VAPOS, VAIW, VAMP and VATM as the explanatory variables. Variables in the long run equation are in their level form while variables in the short run equation are differenced once. The R^{2} and adjusted R^{2} are the coefficient of determination which signifies the proportion of dependent variable that is explained by the independent variables. While the adjusted R^{2} is adjusted for the number of variables in the equation, the R^{2} is not. The Adjusted R^{2} in Table 10 indicates that 68.39% of the variation in real gross domestic product (RGDP) is explained by the financial innovation variables employed in this paper. On the other hand, the R^{2} value shows that 73.38% of the variation in RGDP is explained by at least one of the independent variables used in this paper. The probability value of the model’s F statistic is less than 5% which indicates that the overall model in Table 10 is significant. Finally, the durbin watson stat value shows that the null hypothesis of the serial correlation test is not rejected (that is there is no evidence of firstdegree serial correlation) because the DW value is within 1.55 and 2.45.
4.2.1. Short Run Model Interpretation
Financial innovation proxies such as value of cheque transactions (VACT), value of ATM transactions (VATM) and value of mobile payment transactions (VAMP) are all positively related to Nigeria’s economy (RGDP) while value of internet (Web) transactions (VAIW) and value of POS transactions (VAPOS) are negatively related to Nigeria’s economy in the short term. Meanwhile, only VAMP has a positive significant impact on Nigeria’s real gross domestic product in the short term at 5% and 10% significance level because the probability value of the coefficient is less than 5% and 10% significance level. On the other hand, VACT and VATM have a positive insignificant impact on the Nigerian economy. Ozurumba and Charles (2019) empirical result from their study on financial innovation and Nigeria’s economic growth also shows that financial innovation variables such as financial transactions via NIBSS and agent banking platforms have a positive insignificant impact of Nigeria’s economy. Finally, VAIW and VAPOS have a negative insignificant impact on the economy of Nigeria in the short run. It is also important to note that the error correction term conforms to the three main conditions that:
The coefficient of the error correction term is negative (0.869761).
The absolute value of the error correction term (ECT) coefficient is less than 1 (0.869761 < 1).
The error correction term is statistically significant because its probability value is less than the chosen level of significance (0.0000 < 5%).
The coefficient of the error correction term (ECT) also reveals the speed of adjustment and the time it takes to adjust from shortrun disequilibrium to long run equilibrium. According to the error correction model, the speed of adjustment is 86.98% which signifies the rate to adjust to long run equilibrium. Furthermore, it will take approximately 13.80 months to adjust to long run equilibrium from short run disequilibrium.
4.2.2. Long Run Model Interpretation
According to the long run result, there is a positive relationship between value of cheque transactions (VACT), value of ATM transactions (VATM), value of mobile payment transactions (VAMP) and the Nigeria’s economy (RGDP). Meanwhile, there is a negative relationship between value of internet (Web) transactions (VAIW), value of POS transactions (VAPOS) and Nigeria’s economic performance (RGDP). It is important to note that according to Table 9, value of mobile payment transactions (VAMP) is the only financial innovation indicator that has a positive significant impact on Nigeria economy in the long run such that if there is a 1% increase in VAMP, Nigeria’s economy will grow significantly by 0.062248%. VAMP has a significant impact because the probability value of its coefficient is less than 10% (0.0632 < 10%). On the other hand, value of cheque transactions (VACT) and value of ATM transactions (VATM) have a positive insignificant effect while value of internet (Web) transactions (VAIW) and value of POS transactions (VAPOS) have a negative insignificant effect on Nigeria’s economy in the long run. We can however conclude that improvement in financial innovation channels such as mobile, cheque and ATM payment platforms in Nigeria will have an overall positive effect on the country’s economy both in the shortrun and long run. This is consistent with the empirical result of Okafor et al. (2017) which concludes that there is a positive long run relationship between financial innovation and Nigeria’s economic growth.
4.2.3 PDL MIDAS Equation
Table 11. MIDAS Polynomial Coefficients
Variable 
Coefficient 
Std. Error 
tStatistic 
Prob. 
C 
17.20092 
0.715758 
24.03176 
0.0000 
Polynomial Coefficients 

LOG(VACT) 

PDL01 
0.053913 
0.071287 
0.756291 
0.4579 
PDL02 
0.031145 
0.028842 
1.079856 
0.2925 
PDL03 
0.002754 
0.002234 
1.233057 
0.2312 
LOG(VATM) 

PDL01 
0.144822 
0.06711 
2.157967 
0.0427 
PDL02 
0.056157 
0.025348 
2.215457 
0.0379 
PDL03 
0.004332 
0.001912 
2.266071 
0.0341 
Log(VAPOS) 

PDL01 
0.026553 
0.033822 
0.785098 
0.4412 
PDL02 
0.004986 
0.012109 
0.411791 
0.6847 
PDL03 
9.10E05 
0.000852 
0.106881 
0.9159 
LOG(VAMP) 

PDL01 
0.013649 
0.015893 
0.858795 
0.4002 
PDL02 
0.009261 
0.005783 
1.601415 
0.1242 
PDL03 
0.00067 
0.000418 
1.603475 
0.1238 
LOG(VAIW) 

PDL01 
0.06164 
0.024629 
2.502772 
0.0207 
PDL02 
0.033628 
0.008924 
3.768242 
0.0011 
PDL03 
0.002891 
0.000676 
4.273863 
0.0003 
Rsquared 0.871865 DurbinWatson stat 2.199127 Adjusted Rsquared 0.871865 
Source: Author’s Computation with the aid of Eviews 10
Table 12. Distribution and Lagged Coefficients of LOG(VACT)
Individual Coefficients 

LOG(VACT) 

Lag 
Coefficient 
Distribution 
0 
0.025523 

1 
0.002642 

2 
0.01473 

3 
0.026593 

4 
0.032948 

5 
0.033793 

6 
0.029129 

7 
0.018956 

8 
0.003275 

9 
0.017916 

10 
0.044616 

11 
0.076825 
Source: Author’s Computation with the aid of Eviews 10
Table 13. Distribution and Lagged Coefficients of LOG (VATM)
LOG(VATM) 

Lag 
Coefficient 
Distribution 
0 
0.092997 

1 
0.049835 

2 
0.015336 

3 
0.0105 

4 
0.027672 

5 
0.03618 

6 
0.036026 

7 
0.027207 

8 
0.009726 

9 
0.016419 

10 
0.051227 

11 
0.094698 
Source: Author’s Computation with the aid of Eviews 10
Table 14. Distribution and Lagged Coefficients of LOG (VAPOS)
LOG(VAPOS) 


Lag 
Coefficient 
Distribution 
0 
0.021658 

1 
0.016945 

2 
0.012414 

3 
0.008065 

4 
0.003898 

5 
8.72E05 

6 
0.00389 

7 
0.007511 

8 
0.01095 

9 
0.014207 

10 
0.017282 

11 
0.020175 
Source: Author’s Computation with the aid of Eviews 10
Table 15. Distribution and Lagged Coefficients of LOG(VAMP)
LOG(VAMP) 

Lag 
Coefficient 
Distribution 
0 
0.005058 

1 
0.002192 

2 
0.008102 

3 
0.012671 

4 
0.015899 

5 
0.017787 

6 
0.018334 

7 
0.017541 

8 
0.015407 

9 
0.011932 

10 
0.007117 

11 
0.000961 
Source: Author’s Computation with the aid of Eviews 10
Table 16. Distribution and Lagged Coefficients of LOG (VAIW)
LOG(VAIW) 

Lag 
Coefficient 
Distribution 
0 
0.030904 

1 
0.00595 

2 
0.013221 

3 
0.02661 

4 
0.034217 

5 
0.036041 

6 
0.032083 

7 
0.022342 

8 
0.006818 

9 
0.014487 

10 
0.041576 

11 
0.074446 
Source: Author’s Computation with the aid of Eviews 10
Table 11 to Table 16 contains the PDL/Almon mixed data sampling (MIDAS) result with quarterly real gross domestic product (RGDP) ranging from Q1’2010 to Q’2020 as the dependent variable and the monthly series for value of cheque transactions (VACT), value of ATM transactions (VATM), value of internet (Web) transactions (VAIW), value of mobile payment transactions (VAMP) and value of POS transactions (VAPOS) ranging from January 2010 to April 2020 as the financial innovation proxies which represents the explanatory variables. The PDA/Almon MIDAS result is divided into three sections which are the quarterly coefficients (table 11), polynomial coefficients (table 11) and the individual lagged coefficients (table 1216).
The Adjusted R^{2} from the MIDAS equation indicates that 87.19% of the variation in RGDP is explained by the financial innovation variables and it shows the model has a high predictive power. The Durbin Watson stat value reveals that the residuals of the MIDAS equation are not serially correlated as it is between 1.55 and 2.45.
According to the result of the individual coefficient, at present period, value of cheque transactions (VACT), value of internet (Web) transactions (VAIW) and value of POS transactions (VAPOS) are positively related to Nigeria’s economic performance while value of ATM transactions (VATM) and value of mobile payment transactions (VAMP) are negatively related to Nigeria’s RGDP.
Over 12 months period, the relationship between VACT, VAMP & VAIW and Nigeria’s economy changed direction from negative to positive impact in the long term (6^{th} to 11^{th} month) but VAPOS and VATM are negatively related to Nigeria’s RGDP in the long term. On the other hand, VATM and VAMP is reversed to a positive relation with Nigeria’s economic performance in the short term (1^{st} to 5^{th} month) while VAIW, VACT and VAPOS negatively impacts Nigeria’s economic growth in the short term (1^{st} to 5^{th} month). It is however important to note that, as the number of lagged period (months) keep increasing, the relationship between financial innovation and the Nigeria’s economy is mixed with a nonlinear relationship.
From table 12 and table 16, an improvement in financial innovation (VACT and VAIW) is first followed by the decline in Nigeria’s economy after which the economy starts recovering and improving till the 11^{th} month. Conversely, from table 13 and table 15, an improvement in VATM and VAMP will first boost the Nigeria’s economy after which the positive effect begins to wane and then translate to a declining economic performance in the short term. Lastly, the VAPOS is almost perfectly linearly and negatively related to the Nigeria’s economy. Boost in the value of POS transactions in Nigeria has a positive but fading impact on Nigeria’s GDP at first, after which the effect turns negative and the negative impact continues to weigh in more on Nigeria’s economic performance till the 11^{th} month. We can however conclude that financial innovation especially the mobile payment platform has both a positive short term and longterm impact on the Nigeria’s economy but could also drive a negative economic performance as POS transactions negatively impacts Nigeria’s economy. Specifically, VAMP has a positive short and longterm effect on the Nigeria’s economy while VAPOS has a negative impact on the Nigeria’s economy both in the short and long term.
4.2.4. Robustness Check
Table 17. Robustness Check of the Relationship between Financial Innovation and the Nigerian Economy in the ARDL and MIDAS Model
Equations 
ARDL 
MIDAS 
Comments 
Short term 
VACT, VAMP and VATM are positively related to RGDP while VAIW and VAPOS are negatively related to Nigeria RGDP 
VAMP and VATM reversed from being negatively to positively related to RGDP between (05^{th} lag periods) while VAIW, VACT and VAPOS turned from being positively to negatively related to Nigeria RGDP. 
VAMP and VATM short term relation with RGDP in the MIDAS and ARDL model is consistent. Also, VACT also shows a consistent result up till the 1^{st} lag period in the short term. VAIW and VAPOS negative relation to RGDP in the ARDL model is also consistent with the MIDAS model in the short term. 
Long Term 
VACT, VAMP and VATM are positively related to RGDP while VAIW and VAPOS are negatively related to Nigeria RGDP 
VACT, VAMP and VAIW turned to being positively related to RGDP between (6^{th}  11^{th}) period while VAPOS and VATM reversed to negative relationship. 
VAMP and VACT long term relation with RGDP in the MIDAS and ARDL model is consistent. Also, VATM also shows a consistent result up till the 8^{th} lag period in the long term. VAPOS negative relation to RGDP in the ARDL model is also consistent with the MIDAS model in the long term. 
Source: Author’s Computation (2020)
4.3. Post Estimation Analysis
For ARDL Model
Figure 3. Jarquebera Normal Distribution Test for the ARDL Model
Source: Author’s Computation with the aid of Eviews 10
Table 18. BreuschGodfrey Serial Correlation LM Test Result for the ARDL Model
BreuschGodfrey Serial Correlation LM Test 

Fstatistic 
0.067826 
Prob. F(1,31) 
0.7963 
Obs*Rsquared 
0.085144 
Prob. ChiSquare(1) 
0.7704 
Source: Author’s Computation with the aid of Eviews 10
Table 19. BreuschPaganGodfrey Heteroskedasticity Test for the ARDL Model
BreuschPaganGodfrey Test 

Fstatistic 
1.772761 
Prob. F(6,32) 
0.1364 
Obs*Rsquared 
9.729351 
Prob. ChiSquare(6) 
0.1365 
Scaled explained SS 
2.39443 
Prob. ChiSquare(6) 
0.8801 
Source: Author’s Computation with the aid of Eviews 10
Table 20. Ramsey RESET Test for the ARDL Model
Ramsey RESET Test 


Value 
df 
Probability 
tstatistic 
0.190609 
31 
0.8501 
Fstatistic 
0.036332 
(1, 31) 
0.8501 
Source: Author’s Computation with the aid of Eviews 10
Figure 4. Cumulative Sum of Square Residual (CUSUM) Stability Test
Source: Author’s Computation with the aid of Eviews 10
Table 18 to 20 as well as figure 3 and 4 shows the residuals post estimation test for the ARDL model.
From figure 4, the residuals from the long run model are normally distributed because the probability value of the Jarquebera statistic is greater than the chosen level of significance (0.217960 > 5%).
The BreuschPaganGodfrey heteroskedasticity test and BreuschGodfrey serial correlation LM test result indicates that residuals are homoscedastic and not serially correlated because the probability values of their test statistic is greater than 5%.
Finally, the Ramsey RESET test shows that there is no misspecification bias in the long run equation as the probability value of the F statistic is greater than the level of significance (0.8501 > 5%) while the cumulative sum of squared residual (CUSUM) test reveals that the long run model is stable over the considered time period in this paper (Q1’2010Q1’2020) as the blue line remains within the boundary of the two bounded red lines between 2010 and 2020.
For MIDAS Model
Figure 5. Jarquebera Normal Distribution Test for the MIDAS Model
Source: Author’s Computation with the aid of Eviews 10
Table 21. Correlogram of Residuals
Source: Author’s Computation with the aid of Eviews 10
Figure 5 and table 21 shows the normal distribution and serial correlation test for the residuals of the MIDAS equation. The Jarque bera test signifies that the residuals for the MIDAS model were normally distributed as the probability value of the Jarquebera statistics is greater than the chosen level of significance (0.906569 > 5%). On the other hand, the residuals are not serially correlated as the probability values of the autocorrelation and partial correlation test coefficients for different lag periods were greater than 5% which is the chosen level of significance. The post estimation results of this paper conclude that the residuals of the MIDAS equation are normally distributed and are not serially correlated.
5. Conclusion and Recommendations
This paper provided empirical evidence on the impact of financial innovation on economic growth in Nigeria between January 2010 and April 2020. The study underscored the crucial role played by financial innovation in the Nigeria’s economy. The financial sector is an intermediary between the deficit and surplus sector, thereby introduction of more innovative system will enhance the transfer of funds to the surplus sector. In addition, there will be easy accessibility and availability of funds to both domestic and foreign investors and this will have a positive impact on foreign portfolio investment (FPI), foreign direct investment (FDI) and domestic investment. Investment expenditure will increase and the process of money creation by the banking sector will be improved which will positively impact the aggregate economy and the financial sector development.
The ARDL Model revealed that mobile payments, ATM transactions and cheques transactions are mostly positively related to Nigeria’s economic performance in the long run and short run but value of mobile payments transactions (VAMP) is positively and significantly related to Nigeria’s economy in the short run and long run. Also, the ARDL bound test confirms that there is presence of long run relationship between financial innovation and Nigeria’s real gross domestic product (RGDP) and the CUSUM test shows that the ARDL model is stable over the considered time period (20102020). The PDL MIDAS equation also confirms positive longterm relationship of value of cheque transactions (VACT) and value of mobile payment transactions (VAMP) with the performance of the Nigeria economy as well as also negative shortterm relationship between value of internet web transactions (VAIW), value of POS transactions (VAPOS) and Nigeria’s RGDP. The study therefore submits that the financial innovative channels in the Nigerian financial system have a pivotal role to play in fostering Nigeria’s economic growth and bringing about a positive change in the welfare of citizens. Thus, the following recommendations are in line with the findings of this study:
Given the positive effect of financial innovation variables such as mobile payment, cheque transactions and ATM on the Nigerian economy, the government should intensify efforts to create new forms of financial innovative structures as well as the develop existing ones especially the POS and internet web structures to ensure easy and secure accessibility to financial services and aid easy private sector access to credit & other forms of funds. This will help boost the financial sector development which will in turn have a multiplier effect on other sectors of the economy. Investment expenditure is also expected to increase which is favourable for Nigeria’s economic growth and development
Financial institutions and government should work with the ICT and other relevant sectors to ensure secure transactions over different platforms such as POS, internet and mobile payments. This will help improve the trust of the people on the use of these channels and indirectly aid financial and economic development of Nigeria.
Financial institutions should sensitise people on the use of these various financial innovation channels in other to improve financial literacy. Improving people’s knowledge on these channels will help ensure more funds are available for financial institutions such as banks for credit creation to the private sector which will improve their capital formation and aid productivity.
References
Alex, B. & Calvin, M. (2016). Financial Innovation and Economic Growth: Evidence from Zimbabwe. Journal of Investment Management and Financial Innovations, vol.13 (2), pp. 6575.
Almon, S. (1965). The distributed lag between capital appropriations and expenditures. Econometrica 33, pp. 178–196.
Ansong, A.; MarfoYiadom, E. & EkowAsmah, E. (2011). The Effects of Financial Innovation on Financial Savings: Evidence from an Economy in Transition. Journal of African Business, 12(1), pp. 93–113. https://doi.org/10.1080/15228916.2011.555271.
Bagehot, W. (1873). Lombard Street: A Description of the Money Market.
Chukwunulu, J. I. (2019). Effect of Financial Innovation on the Nigerian Economy. World Journal of Innovative Research, 6 (5), pp. 1521.
Goodhue, D. L. & Thompson, R. L. (1995). Tasktechnology Fit and individual performance. MIS Quarterly, 19(2), pp. 213236.
Gregorio, J. D. & Guidotti, P. E. (1995). Financial development and economic growth. World Development, 23(3), pp. 433–448. https://doi.org/10.1016/0305750X(94)00132I
Idun, Anthony & Aboagye, Anthony (2014). Bank Competition, Financial Innovations and Economic Growth in Ghana. African Journal of Economic and Management Studies, 5(1), pp. 3051.
Johansen, S. & K. Juselius (1990). Maximum Likelihood Estimation and Inference on Cointegration with Applications to Demand for Money. Oxford Bulletin of Economics and Statistics 52, pp. 169210.
John, S. A. (2019). Electronic Payment Systems (Epayments) and Nigeria Economic Growth. European Business & Management.Vol. 5(6), pp. 7787.
Judge, G.; Griffith W.E.; Hill, R. C.; Lutk epohl H. & Lee, T.C. (1985). The theory and Practice of Econometrics. Second Edition (John Wiley & Sons).
MannahBlankson, T. & Belnye, F. (2004). Financial innovation and the demand for money in Ghana. Accra: Bank of Ghana. Working Paper, pp. 1–23.
Nazir, M. R. & Tan, Y. (2018). Financial innovation and economic growth: Empirical evidence from China, India and Pakistan. International Journal of Finance & Economics. DOI: 10.1002/ijfe.2107. pp. 124.
Ogedebe, P. M. & Jacob, B. P. (2012). EPayments: Prospects and Challenges in Nigeria Public Sector. International Journal of Modern Engineering Research (IJMER), Vol. 2(5), pp. 31043106.
Okafor, I. G.; Chijindu, E. H. & Chikezie, A. K. (2017). Does Financial Innovation Drive Growth? Empirical Analysis of the Nigerian Case. British Journal of Economics, Management & Trade, vol. 16(3), pp. 110.
Ozurumba, O. C. & Charles, O. (2019). The Impact of Financial Innovation on Economic Growth in Nigeria. International Journal of Economics, Commerce and Management, vol. 7 (8), pp. 114.
Puatwoe, J.T. & Piabuo, S.M. (2017). Financial sector development and economic growth: evidence from Cameroon. Financial Innovation.
Qamruzzaman, M. & Jianguo, W. (2017). Financial Innovation and Economic Growth in Bangladesh. Research on Financial Innovation. Springer open.
Ravikumar, T.; Suresha, B.; Sriram, M. & Rajesh, R. (2019). Digital Payments on Economic Growth: Evidence from India. International Journal of Innovative Technology and Exploring Engineering (IJITEE), 8 (12), pp. 553557.
Fair, R. & Jaffee, D. (1972). Method of Analysis for Markets in Disequilibrium. Econometrica, 40, pp. 497514.
Romer, P. (1994). The Origins of Endogenous Growth. The Journal of Economic Perspectives, 8(1), pp. 322. www.jstor.org/stable/2138148
Saraswati, N. & Mukhlis, (2018). The Influence of Debit Card, Credit card, and EMoney Transactions toward Currency Demand in Indonesia. Quantitative Economic Research, 1(2), pp. 8794
Michalopoulos, Stelios & Laeven Luc & Levine Ross (2009). Financial Innovation and Endogenous Growth. NBER Working Papers 15356, National Bureau of Economic Research, Inc.
Tahir, S. H.; Shah, S.; Arif, F.; Ahmad, G.; Aziz, Q. & Ullah, M. R. (2018). Does Financial Innovation improve performance? An analysis of process innovation used in Pakistan. DANS Journal of Innovation Economics & Management, Vol. 3(27), pp. 195214.
Tee, H. & Ong, H. (2016). Cashless Payment and Economic Growth. Research Paper on Financial Innovation Vol. 2(4). Open Access paper.
Thomas, S. (2016). Effect of EMoney to economic Performance (a Comparative Study of Selected countries). International conference of Management Sciences. UMY. Indonesia. http://www.researchgate.net/publication/321069476.
William G. Tomek & Maddala, G. S (1977). Econometrics. New York: McGrawHill Book Co. American Journal of Agricultural Economics, Volume 60, Issue 3, pp. 565–566. https://doi.org/10.2307/1239963.
1 University of Lagos, Nigeria, Correspondig author: adeseteahmed@gmail.com.
2 General Studies Department, Federal Polytechnic, Ogun State, Email: auwal.adamu@federalpolyilaro.edu.ng.
3 University of Lagos, Nigeria, Email: rissydauda@yahoo.com.