This paper analyses the asymmetric responses of manufacturing output to changes in exchange rates and bank credit in Nigeria. The results reveal significant countercyclical effects of exchange rate changes on manufacturing output. Bank credit to the private sector was the only predictor with procyclical effects on Nigeria’s manufacturing output. As these responses are often impacted by behavioural patterns, shifts in policy and economic fluctuations, the study employed the non-linear ARDL method alongside the Wald test, Quandt-Andrews and Zivot-Andrews tests. The results of the phase shift analysis show that in Nigeria, only private sector credit leads the manufacturing output cycle, while changes in the exchange rate, inflation and lending rates lag behind. Regardless of the timeframe, manufacturing output was adversely affected by positive and negative changes or variations in the exchange rate. Shifts in bank credit, whether positive or negative, had a positive and considerable effect on manufacturing production. The Wald test confirms the presence of asymmetry in the effects of the exchange rate and private credit on output. The Quandt-Andrews F-statistics for both the maximum likelihood ratio (LR) and Wald statistics, as well as the Zivot-Andrews intercept and trend test results, show that there were breakpoints in bank credit and exchange rate variations in different years, particularly in 2016 and 2020, which marked periods of economic recession, health pandemics, policy shifts, external shocks and macroeconomic instability, as measured by rising petrol pump prices. Neither model, with or without structural breaks, supports the conventional economic theory that devaluation leads to an expansion in output. This is attributed to the contractionary effect of naira depreciation in the context of significant foreign currency-denominated external debt. The negative output effect of naira devaluation in Nigeria was also explained by the low level of competition among domestic firms. In the short term, the results further account for the inflation-output trade-off in Nigeria’s manufacturing industry, while in the long term, the neutrality principle does not fully apply to Nigeria, whose financial market is still emerging, hampered by structural imbalances in the economy. To minimize arbitrage, the Nigerian government should implement financial policies capable of closing the gap between devaluation and appreciation of the naira exchange rates. Specifically, the monetary authorities should set credible inflation targets, and the rate of interest adjustment should align with these targets. This can be achieved by creating an autonomous central bank responsible for maintaining price stability. The research findings will be valuable to manufacturers, the financial sector and small and medium-sized enterprise (SME) owners, both in and outside Nigeria. SMEs and manufacturing industries can benefit from government funding, aid, or investment tax breaks. Such initiatives may take the form of reinvestment allowances, amortization allowances or cash-based grants.

Deposit Money Banks, Private Sector Credit, Naira depreciation, Manufacturing, GDP/Output, Inflation, lending rate, Nigeria.

The Nigerian banking sector plays a pivotal role in the nation’s economic advancement through its lending operations. By lending to private sector and government entities, banks provide the financing necessary for various activities, including production, consumption and business operations. This mechanism drives economic progress. Abina (2020) corroborates this viewpoint, emphasizing the crucial role of deposit money banks in supporting economic expansion in Nigeria by providing credit facilities in the form of loans and lending. According to the CBN (2023), the banking subsector contributes to economic expansion by enhancing manufacturers’ access to credit through lending to finance investment, improving financial stability, fostering employment generation, broadening revenue sources and augmenting foreign exchange earnings in Nigeria. Loans offered by the banking sector to the wider private sector are expected to stimulate investment and thus foster economic expansion. According to the Central Bank of Nigeria (2024), the sectors in the Nigerian economy that benefit from bank lending include the production sector, general commerce sector, services sector and others. However, despite the financial sector’s supposed contribution to the production sector, the rate at which manufacturing has progressed is not encouraging. According to the National Bureau of Statistics, its contribution to GDP declined from 10.20% in the first quarter of 2022 to 10.13% in the first quarter of 2023. In the second quarter of 2023, its contribution fell by 8.62%, which was less than the decline of 8.65% seen in the second quarter of 2022, and lower than the figure of 10.13% in the first quarter of 2023. The Central Bank of Nigeria’s decision to increase interest rates on credit from 11.5% to 135% in 2022 had an impact on people’s willingness to take out loans. Recently, interest rates have risen as high as 22%. This calls for an empirical investigation into the impact of bank lending on Nigeria’s manufacturing sector.

Performance of the country’s manufacturing sector is crucial for industrial growth. Its importance cannot be overstated, as it drives the overall economic progress by fostering employment and technological innovation, and contributes to creating a better society through industrial development. The manufacturing sector encompasses thirteen distinct activities: oil refinement; cement production; food, beverages, and tobacco processing; manufacturing of textile, apparel, footwear, as well as wood products, paper products, chemical/pharmacological products, and rubber goods (CBN, 2023). There is also electrical and electronic manufacturing, basic metal, iron, and steel production, motor and vehicles assembly, and various industrialized activities. Despite numerous challenges that could potentially affect both its output and the overall economy, the manufacturing sector in Nigeria has served as a significant driver of economic activity. The manufacturing sector appears to be essential to any nation’s economy, given its substantial contribution to economic progress.

Regretfully, in Nigeria this sector has demonstrated low capacity utilization over the past decades, which resulted in low manufacturing output and its rather insignificant contribution to the country’s GDP. This has been linked to the sector’s excessive reliance on imports for most of the materials needed for manufacturing. Hardly any intermediate or capital goods are produced. The unfortunate thing is that Nigeria is now importing grains and food products even though it has the capacity to produce enough of these to become self-sufficient (Ac-Ogbonna, 2021). Foreign workers make up the majority of the workforce in the nation’s industrial sector, which is dominated by modern technological expertise. As of 2020, the manufacturing sector’s contribution to Nigeria’s GDP was still under 10%, and its growth had been continuously stunted (CBN, 2020). Despite all efforts to boost productivity, the sector’s poor performance may be related to two factors: (1) the difficulty of obtaining finance in Nigeria, where banks account for almost 90% of the total financial sector’s assets, and (2) fluctuations in the naira/US dollar exchange rate. An important macroeconomic factor that influences how the domestic economy functions in relation to international trade is the exchange rate. According to real option theory, macroeconomic uncertainty has a significant impact on investment choices. Consequently, the exchange rate volatility as a signal of uncertainty can be used to explain investment decisions. When the currency rate fluctuates, its impact on output becomes unclear because of the risks that are probably due to the increased transaction costs associated with the high volatility of the exchange rate. Umoru et al., (2023) note that fluctuations in the naira/USD exchange rate and excessive devaluation of the local currency were major economic problems affecting the Nigerian economy, particularly its manufacturing sector, when investors’ returns are hindered and manufacturing industries’ potential for expansion restricted.

Building on previous research, this research evaluates the asymmetric the influence of banking sector credit and changes in exchange rates on manufacturing output in Nigeria. It also examines how inflation and banks’ lending rates influence the output of Nigeria’s manufacturing sector. It is hypothesized that the asymmetric impact of these variables on the manufacturing sector is not substantial. The findings of this research are relevant to policymakers, governments, property managers and owners, as well as to academic researchers interested in this field of study and seeking to fill the existing gaps. To maximize production, it is important to consider funding and its accessibility. This is why research into the impact of lending by demand deposit banks on the manufacturing sector’s output in Nigeria is so important. The paper sets out guidelines to help the country’s banks and government provide effective support to the manufacturing sector, which is crucial for economic development. Those in charge of manufacturing enterprises will gain insight into how credit from demand deposit banks could benefit their sector.

This paper is divided into five sections. It begins with an introduction, followed by a literature review covering conceptual issues, theoretical discussions and empirical literature. The third section focuses on the methodology and specification of the model used in the study. The fourth section presents and discusses the outcomes of the analysis. The final section contains the conclusions and recommendations drawn from the study.

The researcher obtained the data from the Central Bank of Nigeria’s Statistical Bulletin (CBN, 2024) and the World Development Indicators which are published by the World Bank (2024). The study examined the exchange rate between the US dollar and the Nigerian naira, the country’s local currency. According to this measurement, an increase in the exchange rate results in the naira depreciating against the US dollar, while a decrease in the exchange rate signifies an increase in the naira’s value relative to the US dollar. The private sector’s credit from demand deposit banks was calculated as a percentage of the country’s GDP. The loan rate was determined using the real interest rate of deposit money banks. The lending rate was calculated by adjusting for inflation using the GDP deflator. The yearly percentage increase in consumer prices was used to calculate the inflation rate. The output of Nigeria’s manufacturing sector was measured as a percentage of its GDP. The sample period is from 1981 to 2024. This encompasses forty-four years. It is within this time frame that estimations were made.

Since most time series models are highly susceptible to structural cracks and unexpected shocks in today’s open economy, tests for structural stability are crucial for this study. Consequently, both a model-based Quandt-Andrews test and a variables-based Zivot-Andrews test were conducted. We tested for structural breaks in the dataset using the variables and models (Andrews, 1993; Zivot & Andrews, 1992). These methods test the null hypothesis of no breakpoint, which is rejected if the probability value of the F-statistic falls below 0.005, and accepted otherwise. Accordingly, this study is based on the financial theory of intermediation, as set out by McKinnon and Shaw (1973), and the model specification of Musa et al., (2023). As elucidated in the theoretical review, the McKinnon–Shaw hypothesis establishes a connection between financial institutions and the performance of the real economy. In particular, the researchers observe that the productivity of the real sector relies heavily on the lending services provided by the financial sector. We therefore hypothesize that the services offered by the deposit money banks influence the outcome of the manufacturing industry in Nigeria. The model specification of Musa et al., (2023) is formulated as given in equation (1).

$M=g\left(CPI,BC\right)$ (1)

where M is the output of the manufacturing sector, CPI is the consumer price index as an indicator of inflation, and BC is commercial bank credit. The model used credit from deposit money banks (i.e. from financial sector) and inflation (a control variable) to analyze the influence of the financial sector on manufacturing. The link between loans from financial institutions and the manufacturing sector can be analyzed by adapting the model described by Musa et al., (2023). They defined the output of the manufacturing sector as a function of the consumer price index proxy and credit from commercial banks. To analyze the effect of loans from demand deposit banks on the productivity of the manufacturing sector, we will use the manufacturing sector’s contribution to GDP as our dependent variable. We will also narrow down bank credit to lending from demand deposit banks to the manufacturing sector and the rate at which they lend. We incorporate the rate of inflation as a control variable. This aligns with the works of Oduor et al., (2021) that identified the influence of inflation on Kenya’s industrial sector’s expansion. To test the hypothesis that there is no relationship between changes in exchange rates, credit from demand deposit banks, and the output of the manufacturing sector, we incorporate the rate at which the banks lend and the amount they lend to the sector as control predictors to help us achieve our study objectives. We specify these as

$MSO=g(CRED,EXRV,INRD, INF)$ (2)

where CRED is credit, proxied by lending by demand deposit banks to the manufacturing sector, EXRV is exchange rate variations (depreciation/appreciation), INRD is the interest rate at which demand deposit banks lend. The econometric form is

$MS{O}_{\mathrm{t}}=a+b_{1}CRE{D}_t+b_{2}EXR{V}_t+b_{3}INR{D}_t+b_{4}IN{F}_t+e_t$ (3)

where α is the constant, the coefficients are β_1, β_2, β₃ and β₄ the error is captured as ε_t. It is expected that the interest rate on demand deposits and inflation will negatively affect the MSO, while demand deposit banks’ credit for manufacturing will have a positive effect. The method used for the empirical analysis is the non-linear auto regressive distributed lag model (NARDL). The NARDL, an extension of the ARDL model, is normally used for analyzing time series data, mostly in the context of co-integration and long-run relations among variables. It allows for nonlinear relations amongst the variables. The ARDL model is typically employed for analyzing time series data, primarily in the context of co-integration and long-run relationships between variables. In this model, the assumed linear relationships between the variables are examined. However, in most real-life scenarios, nonlinearities abound. Therefore, NARDL is useful for accommodating circumstances where linear relations may not accurately represent the data. The relationship between variables is expressed through equations incorporating lagged values of the dependent and explanatory variables, as well as interaction terms between them. To capture the complexity of the relationship, these equations can include non-linear functions, such as polynomials. NARDL models can capture complicated and nonlinear relations that may exist in economic and financial data, and so are more accurate in modeling and forecasting. The general form of a NARDL model is

$$\begin{gathered} Z_t=a+b_1{Z}_{t-1}+b_2{Z}_{t-2}+...+b_p{Z}_{t-p}+g_1{X}_{t-1} \\ +g_2{X}_{t-2}+...+g_q{X}_{t-q}+d_1{X}_{t-1}{Y}_{t-1}+d_2{X}_{t-2}{Y}_{t-2}+...+d_t{X}_{t-t}{Y}_{t-t}+{\in }_t \end{gathered}$$ (4)

where; Z_t is the dependent variable at time t, X_t is the vector of explanatory variable at time t, α is constant term, β₁, β₂, ... β_p are coefficients illustrating the influence of lagged values of the dependent variable, g₁, g₂, ... g_q are coefficients illustrating the influence of lagged values of the independent variable, d₁, d₂, ... d_t are coefficients illustrating the influence of the interaction terms between lagged values of the dependent and explanatory variables, ∈_t is error term at time t. The NARDL model used for the empirical analysis incorporates the variables.

$$\begin{gathered} DMS{O}_t=rMS{O}_{t-1}+a_j^{+}CRE{D}_{t-j}^{+}+a_j^{-}CRE{D}_{t-j}^{-}+b_j^{+}EXR{V}_{t-j}^{+}+ \\ +b_j^{-}EXR{V}_{t-j}^{-}+d_j^{+}INR{D}_{t-j}^{+}+d_j^{-}INR{D}_{t-j}^{-}+c_j^{+}IN{F}_{t-j}^{+}+c_j^{-}IN{F}_{t-j}^{-}+ \\ +\sum _{j=1}^p{e}_{j}DMS{O}_{t-j}+\sum _{j=0}^q(h_j^{+}DCRE{D}_{t-j}^{+}+h_j^{-}DCRE{D}_{t-j}^{-})+ \\ +\sum _{j=0}^y(z_j^{+}DEXR{V}_{t-j}^{+}+z_j^{-}DEXR{V}_{t-j}^{-})+\sum _{j=0}^k(d_j^{+}DINR{D}_{t-j}^{+}+d_j^{-}DINR{D}_{t-j}^{-})+ \\ +\sum _{j=0}^s(h_j^{+}DIN{F}_{t-j}^{+}+hDIN{F}_{t-j}^{-})+qec{m}_{t-1} \end{gathered}$$ (5)

where MSO, CRED, EXRV, INRD, INF are as defined earlier, ecm_t-1 is the one-period error correction. For stationarity, the study used an ADF unit root test. The ideal lag for output of the manufacturing sector is p while the lags for credit facility, exchange rate variation, interest rate and inflation rate are represented q,y,k and s respectively. We tested for the presence of a co-integrating vector between the variables after estimating equation (5). This evaluation is based on the non-linear ARDL bounds test methodology, which uses F-test results with critical values for lower and upper bounds designated as  and , respectively. The underlying hypotheses are given as : Zero co-integration vs. H₁: Co-integration. If the computed F-statistic is greater than the upper critical bounds, indicating co-integration, the null hypothesis is rejected. Conversely, if it is less than the lower critical bound, indicating no co-integration, the null hypothesis is accepted. If the F-statistic lies between the upper and lower boundaries, the outcome is indecisive. Also, to verify an ongoing correlation between the variables, the error correction term must be both significant and negative.

The NARDL model, first presented by Shin et al., (2014), offers a thorough approach to analyzing asymmetric interactions and structural fractures between variables in the short and long term. This justifies its use. It is particularly well-suited to studying the impact of variables such as exchange rates, inflation, credit levels and interest rates on output growth, because responses are frequently impacted by economic shocks, policy changes and behavioural patterns. ARDL and other conventional models assume that changes in independent variables have a symmetrical impact on the response variable. In practice, however, positive and negative changes in a variable for example; exchange rate appreciation or depreciation, or interest rate rises or falls do not always produce the same effect. To illustrate the asymmetry in how businesses react to changes in the economy, an increase in interest rates may deter investment more than a drop does. Understanding this distinction is essential for researching real-world financial behaviors, as it ensures more accurate analysis and better policy recommendations.

NARDL, in contrast to linear models, captures asymmetric dynamics, showing how exchange rate fluctuations, both positive and negative, can differently affect investment choices. Significant structural changes have occurred in Nigeria’s economy over time, caused by economic downturns, as well as by the deregulation of interest and exchange rates, which was mandated by the structural adjustment programmes championed by the International Monetary Fund (IMF). These changes have impacted the nation’s financial state. The NARDL model is a reliable approach for examining Nigeria’s economic equilibrium, since it can take into account policy-driven changes that affect economic connections. This NARDL model is particularly useful for managing variables integrated at various degrees, namely I(0)/I(1), and for dealing with small sample sizes, a common issue in time-series data from developing countries. Its ability to capture both short- and long-term fluctuations makes it ideal for research. Compared to typical linear models, the NARDL technique provides more accurate and realistic insights into short-term and long-term fluctuations. Ultimately, this approach ensures that policy inferences accurately reflect the complex economic dynamics of markets. We carried out tests for co-integration, normality and stability to ensure the strength and effectiveness of the NARDL model. The ARCH-LM and Breusch-Godfrey LM tests were conducted to identify heteroskedasticity and serial correlation in the data. Additionally, we used the Jarque-Bera test statistic to determine whether the series was normal. The accuracy of the model specification was determined using the Ramsey Reset test. The study also conducted a Hodrick-Prescott filtering test to determine whether the positive and negative effects of changes to the variables were countercyclical or procyclical.

Table 1 below contains the descriptive statistics. The results show that the variations between the highest and lowest values between 1981 and 2024 are slim for MSO, INRD and CRED. However, the variation for INF is high. The standard deviation of the dependent variable (MSO) is 0.94, indicating that the variation in the contribution of the manufacturing sector to GDP is not significant. The same applies to INRD and CRED, with values of 6.08 and 1.07, respectively. However, INF’s standard deviation value of 16.1 is high. Negative skewness is found in all the variables except INF, which is skewed positively. In character, INRD is mesokurtic, MSO and CRED are playkurtic, while INF is leptokurtic. Based on the Jarque-Bera values and probability values, we can deduce that all the variables except INF are normally distributed.

It is important to check for the presence of a unit root using ADF (Augmented Dickey-Fuller) unit root tests. This is illustrated in Table 2 below. The results of the above test show that INF is stationary at the level, while INRD, CRED and MSO are stationary after being differenced once. This further justifies maximizing the non-linear ARDL approach to achieve the study’s objective.

From Table 3, it is clear that the optimal lag length for this model, as prescribed by the Akaike Information Criterion (AIC), is 2. We then proceeded to ascertain whether a long-term link exists amongst the variables.

The NARDL bounds test results for cointegration are reported in Tables 4A and 4B below, which show the results for the non-linear ARDL model with and without structural breakpoints, respectively. With F-statistic values of 13.5975 and 9.3674 for the NARDL models with and without structural break points, which are higher than the 1%, 2.5%, 5% and 10% upper and lower bound values, the tables show that there is a non-linear long-term link between output in the manufacturing sector, changes in exchange rates, credit from demand deposit banks, the lending rate and inflation. Therefore, we reject the null hypothesis of the absence of a long-term link between the variables and accept the alternative hypothesis of the presence of a long-term link between the variables.

Tables 5A and 5B show the results of the structural break tests, respectively. According to the Quandt-Andrews F-statistics of the maximum LR statistic and maximum Wald statistic published in Table 5B, the alternative hypothesis of a structural break point in the model was accepted for the NARDL equation. This suggests that the NARDL model experienced a structural break in 2020. The Zivot-Andrews intercept and trend test results in Table 5B show that bank credit and exchange rate variations had breakpoints in the years 2014, 2015, 2016, 2020 and 2022. Inflation and the bank lending rate both had structural breakpoints in 2020, whereas manufacturing output had structural breaks in various years: 2016, 2020 and 2022. All the break dates were found to be significant. The breaks in the yearly periods can be attributed to the epidemic of the Coronavirus (Covid-19), macroeconomic instability (measured by rising petrol pump prices, rising prices for goods and services, and falling purchasing power), and a policy shift in exchange rate management, which involved switching from the Retail Dutch Auction System (RDAS), the Wholesale Dutch Auction System (WDAS) and the Interbank Rate System regimes, to adopting a managed floating exchange rate regime. Taken together, these confirm that Nigeria has experienced several periods of economic disturbance, including recessions, political unrest and external shocks. The 2016 and 2020 recessions were so severe that the country saw negative growth rates for three consecutive quarters in those years. This was caused by a number of factors, including falling oil prices, a weak currency and falling investment.

Long-run analysis: The estimated long-run coefficients for the NARDL model are reported in Table 6, which shows the long-term coefficients of the model’s variables. In the long term, a positive change in CRED in the current period had a significant positive impact on the output of the manufacturing sector, MSO. A negative change in CRED had a positive yet non-significant effect; however, when lagged once, its impact was negative and significant. A positive change in INRD had a positive influence on MSO, while a negative change in INRD had a negative influence, though neither impact was momentous. Positive and negative changes in INF in the previous period had a positive and significant influence on MSO. This means that inflation is one of the variables influencing the productivity and progress of the manufacturing sector of the economy. The results show that the lending rate for demand deposit banks did not significantly affect output.

Short-run analysis: Table 7 presents the short-run results of the model. The long-run estimates are comparable to the short-term estimations. For the short-term estimate, negative coefficients were obtained for both exchange rate appreciation and depreciation. These are -1.81090 and -3.09785, respectively. It indicates that fluctuations in the value of the naira relative to the US dollar have a detrimental effect on Nigeria’s manufacturing sector. While the unfavourable effect of appreciation is significant only at the 10% level, depreciation has a highly considerable harmful effect at the 1% level and a zero probability. The coefficient of the error correction term is negative, with -0.446841 as its value and significant at 1 percent level. This illustrates that when there is a disequilibrium in the system, its rate of return to the equilibrium level is 44.68 percent and 66.72 percent for both models with and without structural breaks. This is otherwise known as the speed of adjustment. This shows that about 44.68 percent of the deviations of the model from its equilibrium value in the previous period were corrected in the current period. The adjusted R-squared is 0.686 when adjustments were made for the degree of freedom. The variables that impacted the dependent variable in the short run were negative change in CRED, Δ(CRED_NEG), positive change in INF, Δ(INF_POS), positive change in INF of the previous period, Δ(INF_POS(-1)), and negative change in INF, Δ(INF_NEG). In the short run, Δ(CRED_NEG), Δ(INF_POS), and Δ(INF_NEG) had advantageous yet non-significant impacts on MSO, while Δ(INF_POS(-1)’s impact on MSO was negative but momentous. This implies that manufacturing output in the short term was insignificantly impacted by the positive coefficient of negative change in demand deposit bank loans. Also, the positive coefficient of the positive change in inflation had no significant influence on manufacturing output. The negative coefficient of the previous period’s positive inflation change had a significant influence on the near-term output of the sector. This suggests that inflation in the previous period had a disadvantageous effect on the current output of the manufacturing sector of the economy. The positive coefficient of negative change in inflation had a non-significant effect on the sector’s output. The R-squared value of 0.685800 suggests that this model generally fits the data well, indicating that approximately 69% of the variations in the dependent variable are explained by the independent variables in the model. Other factors impacting the dependent variable that were not factored into the model amount to 31%. When adjusted, the R value became 0.648835, showing that changes in the manufacturing output can be explained by the independent variables, albeit with 35% remaining unexplained.

The results of the short-run and long-run Wald tests for both models with and without structural breaks are reported in Tables 8A and 8B, respectively, for changes in the exchange rate. The Wald test results for changes in bank credit are reported in Tables 9A and 9B, respectively, for the analysis of asymmetric effects. Significant Wald test statistics show the presence of short- and long-run asymmetries in exchange rate and bank credit fluctuations.

Table 10 reports the results of the cyclical behaviour of manufacturing output in relation to changes in the exchange rate, credit from the banking sector, inflation, and the lending rate. According to the most recent empirical findings, fluctuations in exchange rates have an adverse cyclical effect on industrial production in Nigeria. Unfortunately, a depreciating naira led to a decline in exports and manufacturing production. There is a cyclical interaction between bank credit and manufacturing output, indicating some correlation between their levels and variations. Manufacturing output was positively and significantly impacted by favourable and unfavourable shifts in bank credit. This validates the production theory, according to which manufacturers can increase production if they have sufficient financing. This may signify economic expansion, where credit rises and businesses invest more. In addition, borrowing costs may have been low enough to make industrial expenditure financially feasible. Since companies depend on credit to fund their operations, the cyclical nature of bank credit had a favorable effect on manufacturing production.

Serial correlation Test Results. As part of the tests to ascertain the fitness of this model, we carried out a test to check for the presence or absence of serial correlation. The null hypothesis is that there is no serial correlation in the residuals. The probability values of 0.9798 and 0.9687 shown in Table 11 are both greater than 0.05. Therefore, we accept the null hypothesis instead of rejecting it.

Heteroskedasticity Test Results. We performed a Breusch-Pagan-Godfrey test to check for heteroskedasticity in the model. As can be seen in Table 12, the results had probability values greater than the 5% critical value. We can therefore accept the null hypothesis of homoscedasticity in the model.

Stability Test. To check the stability of the model, we carried out the cumulative sum (CUSUM) and Cumulative Sum of Squares (CUSUM of Squares) tests. The model is found to be within the 5% allowance as shown in Figures 1 and 2 below; hence, we accept the null hypothesis of stability of the model.

Figure 1. 

Stability Test Results (CUSUM). Source: Authors’ Eviews 13 plot

Figure 2. 

Stability Test Results (CUSUM of Squares). Source: Authors’ Eviews 13 plot

Normality Test. A Jarque-Bera value of 0.945 and a probability value of 0.62 at the 5% level of significance suggest that the distribution is normal. Figure 3 below illustrates this. Consequently, we can accept the null hypothesis that the model is normally distributed.

Figure 3. 

Normality Test Results. Source: Authors’ Eviews 13 computation

The results of the diagnostic and structural stability tests for the non-linear ARDL model with structural breaks are presented in Table 13. As the results show, autocorrelation is not an issue because the p-value of 0.6487 is higher than the margin of error of 5%, which prevented us from rejecting the null hypothesis that there is no autocorrelation. Having verified the absence of heteroscedasticity, a probability value of 0.3924 revealed that the residuals had a constant mean and common variance over the sample periods. According to the Ramsey RESET test, the model is appropriately stated, since the p-value of 0.8724 is higher than 5%. Furthermore, the study’s findings demonstrate that the residuals linked to the structurally broken non-linear ARDL model are regularly distributed.

In light of the results of the non-linear ARDL model with structural breakpoints, an increase in the exchange rate amounts to a depreciation of the naira, while a decrease in the exchange rate is equivalent to an appreciation of the naira. The results indicate that exchange rate depreciation had a negative impact on the performance of Nigeria’s manufacturing sector in the short-run. Specifically, a 1% depreciation of the naira exchange rate reduces manufacturing output by 1.2130% in the short term, while depreciation worsens output by about 1.8368% in the long term. Even when the exchange rate appreciates in the initial stage, it has a negligible positive asymmetric effect of 0.0013% on manufacturing output in the long term. In the short term, however, the infinitesimal rise in manufacturing output was just 0.0095%. This indeed explains why the manufacturing sector has been performing very poorly in Nigeria since 1980. Since the exchange rate affects the cost of imported goods and services, its volatility results in costs that reverberate through various sectors of the economy (Mwiya et al., 2024). In the short term, manufacturing output in Nigeria declines quickly due to exchange rate depreciation, while in the long term, the response is low. This could be due to the sensitivity of the manufacturing sector to imports of the raw materials needed for production, and the country’s resulting dependence on imports. Given that the manufacturing sector in Nigeria is driven by imports of major inputs, technology and spare parts, depreciation in the value of the naira against the US dollar stimulates a corresponding rise in the cost of production for manufacturers, resulting in a decline in output.

As seen from the negative coefficients of the non-linear model without structural break points, the variations in the value of the currency impede Nigeria’s manufacturing production. The responsiveness of demand for manufactured products in Nigeria varies significantly in view of exchange rate fluctuations, with some products experiencing poorer demand than others when the cost of production escalates owing to depreciation of the naira exchange rate. Our findings are in line with the long-term findings of Onwuka (2022), who assessed the ARCH/GARCH model and ARDL and discovered that exchange rate volatility had a negative effect on the performance of Nigeria’s manufacturing sector among other explanatory factors. Our findings partly concur with those of Uruakpa et al., (2021), who found that, while strong Naira exchange rates had a positive impact on Nigerian manufacturing output, exchange rate volatility had a significant negative impact on the country’s manufacturing output.

The results of our study corroborate those of Vo et al., (2019), who found that fluctuations in the exchange rate of Vietnam’s local currency had a detrimental effect on output in the nation’s industrial sector. Furthermore, our estimations corroborate those of Adegbemi (2018), whose research revealed a negative correlation between increases in Nigeria’s manufacturing output and fluctuations in the naira’s exchange rate. Our findings are consistent with those of Amadi et al., (2018), who found that fluctuations in exchange rates limit the manufacturing sector’s performance in Nigeria and so have a significant macroeconomic impact on the industry. Similarly, our empirical results align with those of Nwokoro (2017), who employed the ECM as the basis for their research and revealed that exchange rates had a substantial and detrimental impact on Nigerian manufacturing production. Furthermore, our results regarding exchange rates are consistent with those of Ali (2020), who found that exchange rate fluctuations hurt manufacturing production, whereas devaluations limit manufacturing sector output. Our study’s results corroborate those of Akeem (2019), who found that exchange rates significantly impacted Nigerian manufacturing output. Additionally, our research results corroborate those obtained by Boateng (2019), who found that the exchange rate of the Ghanaian currency negatively affected the production growth of manufacturing enterprises in Ghana.

At the same time, the findings of our research run counter to those of Musa (2024), who looked into the asymmetric effects of exchange rates on Vietnam’s manufacturing output and showed that fluctuations in exchange rates could have a disproportionate effect on manufacturing output. The study suggests that, although currency depreciation has a significant positive impact on manufacturing, currency appreciation has a less noticeable effect on the sector.

Our findings also contradict those of Ayobami (2019), who concluded that exchange rate changes had a favourable, albeit negligible, impact on the expansion of manufacturing enterprises in Nigeria. Ayobami’s research was based on a linear ARDL approach. Similarly, we found no evidence to support Abdul-Mumuni’s (2019) assertion that manufacturing enterprises in Ghana performed better when the Ghanaian cedi’s exchange rate fluctuated. However, our findings are corroborated by Buabeng et al., (2019), who found that the Cedi/USD exchange rate had a substantial detrimental impact on Ghana’s manufacturing production. In the Nigerian case, the results of both models, with and without structural break points, do not support the conventional economic theory that devaluation leads to an expansion in output because it enhances local production for export, reduces import costs, and stimulates import-competing sectors of the domestic economy. In the Nigerian context, however, incessant naira devaluations have a contractionary effect on external debt denominated in foreign currencies. As of June 2024, Nigeria’s external debt stood at $42.9 billion. This is money owed to foreign creditors, including the World Bank and other bilateral lenders. As of the 2023 financial year, Nigeria’s debt-to-GDP ratio stood at 42%. This exceeds the benchmark set by the Debt Management Office (CBN, 2024). Accordingly, depreciation increases the amount of government money required to service foreign debt, thereby crowding out domestic investment (CBN, 2023). For example, in the final quarter of 2024, the value of the naira fell by 70% (CBN, 2024). Consequently, in March 2024, the Debt Management Office applied a rate of N1,330.26 to $1 to convert the nation’s external debt to naira. Another reason why naira exchange rate depreciation has a negative effect on manufacturing output in Nigeria is the low profit margin in the sector. This is because the cost of imported inputs required for domestic output is high. In addition, Nigeria’s manufacturing sector has a low level of capacity utilization relative to GDP. Excessive reliance on export-led sectors for the inputs used in manufacturing has resulted in low capacity utilization, low productivity and, consequently, low output growth.

This, in turn, increases import costs, which has led to a reduction in the inflow of foreign capital needed by manufacturing firms in the country. Ac-Ogbonna (2021) points out that a shortage of foreign currency restricts the import of materials. The liberalization of the naira/dollar exchange rate, brought about by the Structural Adjustment Programme (SAP), resulted in fluctuating and depreciating exchange rates over time. In fact, the naira has always been adjusted in relation to the dollar since the International Monetary Fund’s (IMF) par value system failed. Over time, this became a policy issue that had fortified the substantial reliance of domestic firms on imports, eventually leading to Nigeria’s unbalanced payments system and the exhaustion of its external reserves. Nigeria’s manufacturing sector is not strong enough for domestic firms to compete on price with imported goods from other countries. This is because manufacturing paper, textiles, tobacco, beverages (including spirits and alcoholic drinks), soft drinks, leather, coal and plastics dominates the sector. Consumer products account for over 90% of total output, while production of capital and intermediate goods is scarce. Taken together, these findings explain why the depreciation of the naira exchange rate has failed to boost manufacturing output in Nigeria. This finding corroborates the conclusion of Orji and Ezeanyaeji (2022) that exchange rate fluctuations hinder manufacturing output.

The short-term outcomes of the NARDL model with structural breaks shed more light on the asymmetrical impacts of deposit money bank credit on Nigerian manufacturing production. With a coefficient of 0.0014, positive change in bank credit had a negligible but positive asymmetric impact when the structural imbalance in the Nigerian economy over the years was taken into consideration. This suggests that bank credit was not large enough to produce significant output in Nigeria’s manufacturing sector. With a value of -0.5596, the negative adjustment in lending by deposit money banks to the private sector had a comparatively large negative impact. This suggests that reductions in private sector credit have a detrimental effect on production in Nigeria’s manufacturing sector. In the estimated nonlinear ARDL equation, a positive change in private-sector credit had a significant positive impact of 1.0152 on Nigeria’s manufacturing GDP when structural imbalances were not taken into account. The model that used structural break points estimated that bank loans had a positive long-term impact on output growth of 0.0085. With a probability value of 0.0490, this effect is positive and considerable at the 5% level. However, this favourable impact is not significant enough to stimulate Nigeria’s manufacturing sector. In contrast, the p-value for a negative shift in private sector credit was 0.0118. When structural imbalances were not taken into consideration over the long term, positive bank credit had a substantial long-term effect (1.0475%) on industrial growth in the model, whereas negative change in private sector credit had a smaller impact (0.0132%). The study does uphold the presence of asymmetry in the effect of deposit money banks’ credit facilities on the manufacturing sector in Nigeria.

This research is important because it shows that increased access to bank loans stimulates manufacturers in the private sector to invest in new equipment and expand their operations. In the long term, this results in higher manufacturing output levels. This contradicts the theory of long-run neutrality of money, which suggests that variations in the money supply in the financial system have no real influence on long-term output or employment; instead, they affect nominal variables such as wages. This theory suggests that monetary and/or financial policies cannot encourage long-term real output growth, but can only affect price levels. The study shows that positive changes in deposit money banks’ credit have a profound long-term impact on manufacturing output, challenging the traditional neutrality assumption. The finding that improvements in bank credit stimulate output growth is consistent with the theory that growth in the financial sector alleviates the constraints manufacturers face in Nigeria. This suggests that monetary interventions such as credit expansion may have long-lasting effects on real output, especially in Nigeria where there are flaws in the financial market. Therefore, structural problems could allow changes in the money supply to affect actual output. The concept of long-term neutrality may not fully apply to an economy like Nigeria’s, where the financial sector is still developing, even though it is applicable in highly developed nations. This underscores the importance of making financial resources accessible for business growth, loan guarantees, credit easing measures or targeted subsidies, as these contribute considerably to the overall growth of the industrial sector. Our findings contradict those of Ilgar and Rukhsara (2023), whose research showed that manufacturing output was not influenced by demand deposit bank activities. In contrast, the results of the present study align with theoretical postulations and corroborate the findings of prior studies, including those by Sigah (2022), Igbinedion (2023), Rubén and Aurea (2023), Njoku et al., (2023) and Ahmad et al., (2024), which all support the finding that private sector credit has a significant positive impact on manufacturing output.

Rubén and Aurea (2023) carried out a similar study in Mexico and obtained results that align with those of this study. Credit from banks had a significant impact on the expansion of the manufacturing sector and other industries. Njoku et al., (2023) obtained similar results in Nigeria. They identified a strong and significant correlation between bank credit and growth in the manufacturing sector. Adesola (2023), in his study, found that the effect of bank credit on economic progress was insignificant in the short term, but significant in the long term. These findings illustrate the relationship between the financial sector and economic advancement. Magaji et al., (2023) found that credit from commercial banks had a positive impact on Nigeria’s output growth in both the short and long term. Ekwunife et al., (2023) obtained very similar results. Their findings emphasize the important role of the financial sector in advancing the manufacturing sector, which was positively impacted by the financial sector in both the short and long term.

In the short term, the equation that allowed for structural break points estimated an insignificant positive coefficient of 0.0246 for a negative change in the lending rate, while a significant negative coefficient of -1.3209 was found for a positive change in the lending rate of deposit money banks. In the equation without structural breakpoints, a positive change in the loan rate had a negative impact on manufacturing GDP, with a coefficient of -2.3287. Conversely, a negative change in the lending rate increased manufacturing output by 0.0148. These estimates at level are comparable to the asymmetric results at one-period lag. In the long term, the model controlling structural imbalances showed positive and negative responses of 0.0597% and 1.0143%, respectively, to an increase in lending rates. Similarly, when structural imbalances are permitted in the model, output decreases by 0.0215% in response to an increase in the lending rate and increases by 0.0103% in response to a decrease. Consequently, Nigeria’s manufacturing production growth was affected asymmetrically by the loan rate. These outcomes contradict those of Igbinedion (2023), who found that interest rates do not affect Nigerian manufacturing production. The outcome is also consistent with economic theory. For the NARDL model with structural breaks, changes in inflation had a long-term impact on manufacturing firm output, reducing it by 0.0185% and 0.0001%, respectively, following positive and negative changes. For the NADRL model without structural breakpoints, a positive change in long-term inflation impacted output by -0.0192%, while a negative change in inflation caused a 0.0015% decline in manufacturing output. In the model that accounted for structural imbalances, a positive long-run change in inflation reduced manufacturing GDP by 0.0185%, while a negative change in inflation also resulted in a 0.0001% drop in long-run output growth. This illustrates the trade-off between inflation and output growth. In the long run, the trade-off is expected to disappear, meaning that sustained inflation does not lead to increased output. The short-term results of this study’s empirical analysis show that, according to the NARDL model with structural break points, both positive and negative changes in inflation caused a significant drop in manufacturing output of 1.2153% and 0.0266%, respectively, in Nigeria. In the model without structural imbalances, a positive change in inflation had a considerable effect, reducing manufacturing GDP by 0.0175%, while a negative change in inflation caused an insignificant drop in manufacturing output of 0.0146%. These suggest that the manufacturing sector’s productivity is not asymmetrically affected by changes in inflation. In agreement with the Phillips curve, which shows an inverse relationship between inflation and output (or unemployment), these short-term findings help clarify the relationship between inflation and output. The study found that manufacturing output decreased when inflation increased, indicating a short-term inverse relationship between the two. This strongly suggests the presence of a short-term trade-off between inflation and output growth within Nigeria’s industrial sector. Sticky wage levels and pricing may help to explain this. This finding is important because it implies that Nigeria’s anti-inflationary policies have not been effective in boosting the productivity of its manufacturing sector. Therefore, the favourable impact of inflation control is negligible, as evidenced by the fact that prices in Nigeria rarely decline because production costs remain high. Since prices in Nigeria rarely fall, real wages tend to decrease. This prompts the labour force to demand higher wages in order to offset the rising prices of goods and services.

To some extent, the industrial sector needs policies that encourage low inflation. This would boost productivity and competitiveness in the sector. These findings call for policies that could increase cost-effectiveness and competitiveness, leading to higher demand and increased manufacturing production in Nigeria. The negative effect of inflation fluctuations on output suggests that inflation has serious adverse effects on Nigeria’s manufacturing sector, most likely due to structural issues such as inflexibility in price adjustments or weak financial markets. In summary, the results show that an increase in inflation leads to a reduction in manufacturing output. This lends support to the traditional Phillips curve theory, emphasizing the importance of balanced monetary and fiscal policies to promote production growth and price stability. This finding aligns with the results of of Oduor et al., (2021) who revealed a harmful effect of inflation on the growth of manufacturing sector in Kenya. The results do not align with those reported by Musa et al., (2023), Mwinuka and Mwangoka (2023), or Ademola (2023). Musa et al., (2023) found that, in the short term, inflation had a significant positive impact on the manufacturing sector’s output. Mwinuka and Mwangoka (2023) also observed that inflation had a positive impact on output and productivity in the manufacturing sector. Mwinuka and Mwangoka (2023) also observed that inflation had a positive impact on output and productivity in the manufacturing sector. Ademola (2023) further confirmed that inflation contributed favourably to manufacturing productivity in Nigeria.

To further verify the nonlinear effects of the key variables in the study, the Wald test of asymmetry was conducted. Wald tests for short-run asymmetry show that the null hypothesis of no asymmetry in the short or long run is rejected for the nonlinear ARDL model, whether or not structural breaks are present. This indicates that depreciation and appreciation of the official naira exchange rates have diverse effects on output of the manufacturing sector in Nigeria. The existence of short-term asymmetry lends support to calls for the Nigerian government and central banking authorities to narrow the gap between the devaluation and appreciation of the naira exchange rate. In the case of bank credit, the non-linear ARDL model without structural breaks rejected the Wald test hypothesis for both short- and long-run asymmetries. However, the model with structural breaks confirmed the presence of asymmetry for both short- and long-term periods of investigation. This can be inferred from the probability values being less than 0.05. This suggests that changes in banking lending only affected the Nigerian manufacturing sector asymmetrically when there were structural imbalances in the country’s economy. Ultimately, as indicated by the likelihood values of the Wald statistics, both banking sector credit and the exchange rate dynamically influence the productivity of Nigeria’s manufacturing industry.

An empirical discussion of the cyclical behaviour of manufacturing output as a percentage of GDP in Nigeria revealed that the exchange rate is the most volatile predictor in the study, with coefficients of 99.674 and 94.513 obtained in analyses with and without structural breaks, respectively. This demonstrates the highly volatile and persistent nature of the naira exchange rate in relation to the U. S. dollar. The high level of economic fluctuation in Nigeria has been attributed to significant exchange rate fluctuations. (Oladipo et al., 2023). We also obtained significant countercyclical output effects from exchange rate changes, as indicated by contemporaneous correlation coefficients of -0.9728 and -0.8534 in models with and without structural breaks, respectively. This explains why manufacturing output in Nigeria fluctuates so much in response to changes in the exchange rate. Taking into account the positive contemporaneous correlation coefficients of 0.6692 and 0.6520 for both models with and without structural imbalances, the bank credit to the private sector was the only predictor that had procyclical effect on Nigeria’s manufacturing production. This suggests that when financial resources are made available to firms and private investors, industrial production rises. Nigerian manufacturing output exhibited a countercyclical nexus with both inflation and the lending rate. It also implies that when prices increase, industrial output tends to increase as well. For models with and without structural imbalances, the inflation-related contemporaneous correlation coefficients are -0.7345 and -0.5639, respectively. Similarly, for models with and without structural imbalances, the contemporaneous correlation scores for the loan rate were -0.4568 and -0.4076 correspondingly. The results of the phase shift analysis show that in Nigeria, only private sector credit leads the manufacturing output cycle, while changes in the exchange rate and other predictors, such as inflation and the lending rate, are lagging predictors. This contradicts the findings of Oladipo et al., (2023), who reported that inflation is the only macroeconomic variable leading the cycle of output growth in Nigeria’s manufacturing industry.

This study analyzed the impact of credit from demand deposit banks on the output and development of the manufacturing sector of the Nigerian economy between 1981 and 2024. To address the nonlinearities in the Nigerian manufacturing industry, a non-linear ARDL approach was adopted. Based on the Quandt-Andrews and Zivot-Andrews test methods, we found significant evidence of structural breaks in the variables and model specification. We also carried out a NARDL bounds test to ascertain a long-term link between the variables. The results showed that fluctuations in the value of the naira against the US dollar negatively impact Nigeria’s manufacturing sector, as indicated by the respective coefficients for exchange rate appreciation and depreciation in both the model with structural breaks and the model without structural breaks. Overall, this study empirically established that a strong link exists between the financial and manufacturing sectors of the Nigerian economy, with the former influencing the latter. Variations in bank lending and the exchange rate have a cyclical effect on Nigeria’s industrial production. Changes in the currency rate have negative cyclical impact on manufacturing output. Both the positive change in bank credit and the negative shift had a large and beneficial impact on manufacturing production. However, the bank’s lending rate did not seem to have a significant impact on manufacturing output. The research findings derived from models with and without structural breakpoints do not support the conventional economic theory that devaluation leads to an expansion in Nigeria’s output. This was attributed to a number of factors. These include low capacity utilization as a percentage of GDP, the dominance of consumer items rather than capital and intermediate goods, and the lack of competitiveness of domestic firms. Bank loans to the private sector had a procyclical impact on Nigerian manufacturing production, whereas exchange rate fluctuations had a significant countercyclical effect on output growth. A long-term link was established and positive and negative changes in inflation had a significant adverse effect on manufacturing output. A positive change in deposit bank credit had a positive effect on output. However, a negative change had an insignificant influence. A positive change in the lending rate had a significant negative impact. A negative change in the rate had a positive impact on the sector, though this was not significant.

The study shows that changes in interest rates have a limited immediate effect on manufacturing output. This indicates that monetary tools such as interest rates might not be sufficient to stimulate the industry. Therefore, strategies aimed at boosting economic growth that do not involve monetary policy, such as sector-specific subsidies or tailored fiscal policies, should be considered. The detrimental effect of a decrease in inflation on manufacturing production highlights the importance of anti-inflationary measures that will truly benefit the industry. Furthermore, measures that support price stability are advantageous, as demonstrated by the finding that lower inflation favourably impacts industrial production. According to the study’s findings, structural problems and flaws in the financial system may prevent money from being completely neutral in the long term in Nigeria. The Nigerian manufacturing sector is also subject to the inflation-output trade-off. The results suggest that, to achieve sustainable growth in the sector, policy interventions should focus on improving access to loans, managing inflation, and implementing structural reforms, rather than relying solely on traditional monetary measures. The Nigerian government should execute financial polices capable of closing the gap between devaluation and appreciation of the naira exchange rates to minimize arbitrage. Specifically, credible inflation targets should be implemented by the monetary authorities, and the rate of interest adjustment should align with such targets. Creating an independent central bank with the responsibility of maintaining price stability will accomplish this. Additionally, the monetary authorities ought to think about influencing aggregate demand and reducing inflation through interest rate changes. The floating exchange rate regime can preserve price stability and lessen the effects of external shocks. The monetary and fiscal authorities of the Nigerian economy should jointly implement policies that narrow the imbalance between the depreciation and appreciation of the naira exchange rates. Specifically, the monetary authorities could implement a managed floating exchange rate system with some intervention. To increase industrial productivity and competitiveness, it is necessary to prioritize public infrastructure investments. Measures to increase productivity and competitiveness in the manufacturing sector should be adopted, such as facilitating loan availability, lowering regulatory costs, and encouraging technological advancement. In this instance, support for local manufacturers should take the form of measures such as subsidies or discounted loan facilities. The government should invest in infrastructure initiatives that improve industrial logistics, reduce transport costs and open up markets. They should encourage market competition in the manufacturing sector by implementing policies including deregulation, promotion of foreign investment, and the privatization of state-owned businesses. This would lower expenses and increase efficiency. Monetary and financial policies that enhance deflationary output growth should be implemented. Fiscal policies aimed at improving the competitiveness of export-led industries should be implemented. These research findings should prove valuable to manufacturers, financial sector executives and small-to-medium enterprise owners, both in and outside Nigeria. The country needs to develop technological innovations and infrastructure to support SMEs in their operations. To lower the cost of imported goods, small and medium-sized enterprises should be offered import duty reductions and manufacturing subsidies. The government may offer reinvestment allowances or incentives to SMEs and manufacturers who are building up manufacturing capacity or branching out into related products. SMEs and local industries can benefit from government funding, aid, investment tax breaks, amortization allowances, or cash-based grants. Future researchers need to evaluate the non-linear effects of bank loan financing and currency rate fluctuations on industrial output in developing financial markets. They should also use the cross-sectional augmented ARDL method to evaluate regime shifts and threshold effects of the variables for each country, and estimate Markov switching and threshold autoregressive models.