See discussions, stats, and author profiles for this publication at: https://www.researchgate.

net/publication/374525058

Effect of Exchange Rate on Domestic Price Level in Nigeria

Article in Asian Research Journal of Arts & Social Sciences · October 2023
DOI: 10.9734/arjass/2023/v21i3469

CITATION READS

1 318

7 authors, including:

Nwokoye Mathew Ignatius Okoye Machi

Nnamdi Azikiwe University, Awka Nnamdi Azikiwe University, Awka
12 PUBLICATIONS 7 CITATIONS 36 PUBLICATIONS 8 CITATIONS

SEE PROFILE SEE PROFILE

Azuka Nwokocha Onwuka Irene Nkechi

Nnamdi Azikiwe University, Awka Nnamdi Azikiwe University, Awka
7 PUBLICATIONS 2 CITATIONS 15 PUBLICATIONS 17 CITATIONS

SEE PROFILE SEE PROFILE

All content following this page was uploaded by Nwokoye Mathew on 09 October 2023.

The user has requested enhancement of the downloaded file.

 Asian Research Journal of Arts & Social Sciences

Volume 21, Issue 3, Page 26-38, 2023; Article no.ARJASS.107165

ISSN: 2456-4761

Effect of Exchange Rate on Domestic

Price Level in Nigeria
Nwokoye Matthew Okechukwu a*, Machi Ignatius Okoye a,
Nwokocha Azuka Florence a, Eze Chioma Confidence a,
Onwuka Irene Nkechi a, Okechukwu Stanley Chidinna a
and Ekwugha Juliet Chika a
a Nnamdi Azikiwe University, Awka, Nigeria.

Authors’ contributions

This work was carried out in collaboration among all authors. All authors read and approved the final
manuscript.

Article Information
DOI: 10.9734/ARJASS/2023/v21i3469

Open Peer Review History:

This journal follows the Advanced Open Peer Review policy. Identity of the Reviewers, Editor(s) and additional Reviewers,
peer review comments, different versions of the manuscript, comments of the editors, etc are available here:
https://www.sdiarticle5.com/review-history/107165

Received: 20/07/2023
Original Research Article Accepted: 27/09/2023
Published: 06/10/2023

ABSTRACT
Nigeria experienced significant depreciation of her currency during the specified period of this study,
mainly due to factors such as falling oil prices, external economic conditions such as slowdown in
global economic growth, trade tensions which led to capital outflows from emerging markets
including markets in Nigeria, tightening of global monetary policy which affected capital flows to
these emerging markets, coupled with domestic macroeconomic challenges like concerns over
policy consistency, governance issues and security challenges, capital flight and weakened
investors’’ confidence. These conditions collectively created an adverse environment for the
Nigerian economy, leading to the depreciation of Naira. Consequently, this study set out to achieve
the main objective of investigating on a more recent basis, effects of exchange rate on domestic
price level in Nigeria. The period covered was from January 2015 through December 2022. The
study adopted the Autoregressive Distributed Lag (ARDL) model and the variables included
consumer price inflation, nominal exchange rate, import prices, international crude oil prices and
_____________________________________________________________________________________________________

*Corresponding author: Email: [email protected];

Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

real output growth. We found insignificant positive impact of the nominal exchange rate on
consumer price inflation in Nigeria. The import prices also proved a significant effect on consumer
price inflation in Nigeria. the study recommends that Governments at all levels in the country should
encourage and support the innovative ideas of business firms and individuals. This will support local
production, hence, there can be substitution of imported goods for domestically produced goods
and the exchange rate will be stable.

Keywords: Exchange rate; domestic price level; real output growth; Nigeria; ARDL.

JEL: C01, E31, F31, O4

1. INTRODUCTION as a result of more robust institutional

frameworks, which are often characterized by
Domestic price level is the average of the current high productivity and low levels of corruption.
prices across the entire range of goods and This stability is further supported by sound and
services within a country's market. It is an effective monetary and fiscal policies, and a
important economic indicator that reflects the more diversified economic base. However, these
overall level of inflation and the purchasing countries are not immune to inflationary
power of consumers, and is a crucial pressures, with factors such as global economic
macroeconomic indicator that provides insights shocks, supply chain disruptions, and shifts in
into the overall health of an economy. Mankiw [1] consumer behavior all potentially contributing to
defined domestic price level as the average level inflationary pressures.
of prices in the economy, measured by a price
index such as the consumer price index (CPI), In Nigeria, the formulation and implementation of
the gross domestic product deflator (GDP monetary policy by the Central Bank of Nigeria
deflator) or the producer price index (PPI). (CBN) is aimed at maintaining price stability
Similarly, in their book "Principles of which is consistent with the achievement of
Macroeconomics”, Mankiw and Taylor [2] sustainable economic growth. However, high and
explained that the domestic price level is affected volatile domestic price level in Nigeria has been
by various factors, including changes in the a major impediment to overall economic growth,
money supply, shifts in aggregate demand and as it leads to reduced purchasing power of
supply, and fluctuations in international trade. especially the vulnerable segment of the
population, discourages savings, induces
In developed countries, domestic price levels are uncertainty for investment, and raises the cost of
typically more stable than in developing countries, doing business [3].

INFLATION RATE (YEAR-ON-YEAR CHANGE, %)

24.00
22.00
20.00
18.00
INFLATION RATE

16.00
14.00
12.00
10.00
8.00
6.00
4.00
2.00
0.00
2015M1
2015M5
2015M9
2016M1
2016M5
2016M9
2017M1
2017M5
2017M9
2018M1
2018M5
2018M9
2019M1
2019M5
2019M9
2020M1
2020M5
2020M9
2021M1
2021M5
2021M9
2022M1
2022M5
2022M9

PERIOD (QUATERLY)
Fig. 1. Inflation rate trend in Nigeria (2015–2021)
Source: Researcher's Compilation (2023), using data from the Central Bank of Nigeria

27
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

According to the National Bureau of Statistics [4], authors such as Ijeoma [5], who found a positive
Nigeria's inflation rate rose to a four-year high of relationship between exchange rate and inflation
18.17% in March 2021, up from 17.33% in in Nigeria, while Nwankwo and Okafor [6], found
February 2021. The rate grew faster than usual a negative relationship. According to Dornbusch
for the 10th straight month to 21.47 percent in and Fischer [7], exchange rate movements can
November of 2022 from 21.09 percent in October affect a country's trade balance, inflation, and
and above market estimates of 21.15 percent. interest rates. As explained by Mordi and
The high inflation rate has been attributed to Adeleke [8], it is a key determinant of inflation in
several factors, including insecurity, the COVID- Nigeria, and its effects can be transmitted
19 pandemic, and the naira's devaluation. High through both imported goods and can also
inflation has also had adverse effects on directly affect the price of goods sold in the
standard of living, as it leads to reduced domestic markets.
disposable income, lower purchasing power and
a decline in the welfare of the Nigerian populace. It is against this backdrop that this study
examines the effect of exchange rate on
In Nigeria, the exchange rate has experienced domestic price level in Nigeria. The rest of the
fluctuations over time, reflecting the dynamics of paper is structured as follows. Section two
the economy and external factors. While there presents the review of empirical literature, while
have been periods of stability and managed section three expresses the data and
exchange rate regimes, there have also been methodology. While section four presents the
episodes of volatility and significant depreciations. analysis and interpretation of result, section five
These fluctuations have been influenced by shows the conclusion and policy
various factors, such as global oil price shocks, recommendations.
economic crises, and policy adjustments. As a
result, the Nigeria exchange rate has exhibited a 2. REVIEW OF EMPIRICAL LITERATURE
mixed pattern, characterized by both
appreciations and depreciations, which can have In Musa's [9] study, the aim was to assess the
varying impacts on the domestic price level. impact of exchange rate volatility on inflation in
These challenges have contributed to foreign Nigeria using annual time series data from 1986
exchange shortages, weakening the naira and to 2019. The study employed the generalized
driving up inflation rates. autoregressive conditional heteroskedasticity
(GARCH) and vector error correction model
Policymakers in Nigeria have responded to (VECM) to examine the long-run effects of
exchange rate instability and high inflation with exchange rate volatility on inflation. The variables
various programs and policies. CBN has included in the model were consumer price index,
increased interest rates and implemented foreign nominal exchange rate, money supply, import,
exchange controls to stabilize the exchange rate and export. The findings indicated that both
and reduce inflation rates. With some success money supply and nominal exchange rate had a
however, these policies have also had negative positive and significant impact on the consumer
effects on the economy. Nigeria has also price index, suggesting that inflation in Nigeria is
employed monetary policy interventions and influenced by fluctuations in the exchange rate
diversification strategies to address the issue. and an increase in money supply.
The country is promoting the growth of non-oil
sectors such as agriculture, manufacturing, and Ari et al. [10] conducted a study on the exchange
services to reduce reliance on foreign exchange. rate pass-through (ERPT) to inflation in
They have prioritized infrastructure development Mozambique by analyzing monthly data from
to attract foreign investment and improve the 2001 to 2019. Their findings indicate that the
country's competitiveness. They have embarked ERPT is asymmetric, significant, and rapid, with
on projects to improve the country's transport, approximately 50 percent of exchange rate
power, and communication infrastructure to variations transmitting to prices within a span of
reduce business costs and improve the overall less than six months. The authors employed the
economic environment. Efforts to maintain price auto-regression distributed lag (ARDL) model to
stability and manage exchange rate fluctuations analyze various variables, including Mozambique
in Nigeria have been ongoing, but challenges CPI, bilateral exchange rates of the Mozambican
persist. Metical (MZN) against the South African Rand
(ZAR) and the United States Dollar (USD),
The relationship between exchange rate and nominal effective exchange rate, import price
inflation has also been explored by several index, real money supply (M3), and rainfall index.

28
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

In their study, Gbalam and Dumani [11], Hernán and Norberto [15] conducted a study
examined the relationship between the general focusing on the nature of exchange rate pass-
price level and foreign exchange rate in Nigeria through to prices across the distribution chain.
using annual historical data from 1990 to 2018. They employed a logistic smooth transition
The variables included in the analysis were the autoregressive vector model and utilized
general price level (proxied by CPI), exchange Bayesian methods to estimate the model using
rate, and real interest rate. The findings indicated data from the Colombian economy and its main
that the foreign exchange rate had a positive but trading partners. The study covered the period
insignificant impact on the level of inflation in from January 2002 to May 2015 and included
Nigeria. This suggests that the changes in the variables such as CPI inflation, real exchange
general price level in the country are not primarily rate, output gap, economic openness, commodity
influenced by imported inflation. prices and interbank interest rate (IBR). The
main finding of their research indicates that
Usman and Mohammad [12], conducted a study exchange rate pass-through is non-linear and
to examine the extent of exchange rate pass- varies depending on the state of the economy
through to consumer price inflation in Nigeria, and the specific shocks experienced.
considering both the short-run and long-run
dynamics. They utilized historical annual data Abiodun et al. [16] conducted a study to examine
covering the period from 1960 to 2015. The study the transmission of exchange rate effects on
employed the Cointegrated Autoregressive import prices and consumer prices at the
Model proposed by Johansen (1988, 1995) to aggregate level in Nigeria. They utilized quarterly
analyze the relationships. The findings indicated historical data from 1995 to 2015 and included
that the exchange rate had a positive and variables such as inflation (CPI), world import
significant effect on the consumer price index prices (represented by the US producer price
(CPI) in both the short run and long run. index), nominal exchange rate, nominal effective
Additionally, the study found that the import price exchange rate, real output (proxied by Real
index and trade openness index had positive and GDP), and oil prices. The study established two
significant effects on consumer price inflation, equations, namely the baseline model for
although some statistical flaws were observed in showing transmission of exchange rate effect to
the short-run impact of the import price index. import prices and alternative model indicating the
transmission to consumer prices. They applied
In their empirical study, Tela et al. [13] Johansen cointegration and vector error
investigated the impact of domestic price level correction techniques to analyze the collected
and monetary policy on exchange rate data. The results indicated that the effect of
fluctuations. They utilized time series annual data exchange rate pass-through is higher in imports
covering the period from 1987 to 2014. The study compared to aggregate consumer prices in
employed the Ordinary Least Square technique Nigeria.
to analyze variables such as real exchange rate,
crude oil prices, consumer price index, domestic In a study by Victor [17], annual data from 1974
interest rate, terms of trade, monetary policy rate, to 2013 was utilized to investigate the factors
foreign exchange reserve, and foreign interest contributing to inflation and their respective
rate. The results indicated a significant magnitudes in Nigeria. The study aimed to
relationship between commodity prices, nominal identify both traditional and institutional variables
interest rate, crude oil prices, monetary policy, associated with the inflation phenomenon and
and the exchange rate. determine their impact on the general price level.
The variables included in the analysis were the
Helmy, Fayed, and Hussien [14] conducted a consumer price index (CPI), lagged inflation, real
study in Egypt to analyze exchange rate pass- wages, output gap, real profits, real effective
through (ERPT). Their research findings revealed exchange rate, crude oil index, real broad money
that the pass-through effect was high but supply, and dummies for institutional density and
incomplete, and it occurred at a slow pace product market density. The historical data
across three categories of prices: consumer price collected was analyzed using the autoregressive
index, producer price index, and import prices. distributed lag (ARDL) model. The findings
The authors further explained while reporting indicated a short-run relationship where 60% of
their findings that the consumer basket in Egypt disequilibrium errors from the previous year's
is significantly influenced by subsidized shock converged back to the long-run equilibrium
commodities and goods with regulated prices. in the current year.

29
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

Lariau et al. [18] conducted a study comparing other relevant variables in the model. We have
the magnitude of exchange rate pass-through incorporated international crude oil prices and
(ERPT) in Angola and Nigeria. The findings real output growth as relevant control variables.
revealed that in Angola, the long-run ERPT was Including international crude oil prices in our
substantial but had been decreasing in recent model is crucial for analyzing the relationship
years due to de-dollarization efforts. On the other between exchange rate and domestic prices,
hand, in Nigeria, the long-run ERPT was found to particularly in the context of an oil-rich economy
be statistically insignificant. However, in the short like Nigeria. This variable allows us to account
run, there was a significant estimate of ERPT for for global shocks that can have a substantial
nonfood prices. impact on Nigeria's economy, considering that oil
receipts play a significant role in the country's
Cheikh NB, et al [19] conducted a study on revenue.Our empirical model takes the following
Recent developments in exchange rate pass- functional form:
through: What have we learned from uncertain
times?. The study reveals the asymmetric effect CPI = f (NER, USWPI, OILP, RGDPG) (1)
of the exchanges rate on prices, depending on
whether the threshold variable is above or below Where;
a certain threshold.
CPI = Domestic consumer prices; NER =
Hassan TA, et al [20]. Conducted a research on Nominal exchange rate; USWPI = Import
A risk-based theory of exchange rate stabilization price index; OILP = International crude oil
and they find a small economy stabilizing its prices; RGDPG = Real output growth. The
bilateral exchange rate relative to a larger econometric representation of equation (2) is
economy can increase domestic capital expressed in the form:
accumulation, domestic wages, and even its
share in world wealth. In the absence of policy CPI = 𝛽0 + 𝛽1 NER + 𝛽2 USWPl + 𝛽3 OILP +
coordination, small countries optimally choose to 𝛽4 RGDPG + εt (2)
stabilize their exchange rates relative to the
currency of the largest economy in the world,
To facilitate the interpretation of estimated
which endogenously emerges as the world’s
coefficients as elasticities and to address the
“anchor currency”, Larger economies instead
issue of heteroscedasticity, the econometric
optimally choose to float their exchange rates.
model is transformed into natural logarithmic
The model therefore predicts an equilibrium
form, allowing for a long-run regression analysis.
pattern of exchange rate arrangements that is
The equation can be represented as follows:
remarkably similar to the one in the data.

Andriyanto RW,[21] investigate on Analysis of Ln(CPI) = 𝛽0 + 𝛽1 Ln(NER) + 𝛽2 Ln(USWPl) +

Oil Price and Exchange Rate in Indonesia, the 𝛽3 Ln(OILP) + 𝛽4 Ln(RGDP) + εt (3)
study used the Error Correction Model (ECM)
and the study's results indicate a significant Where;
negative effect between the variables of foreign
exchange reserves, relative GDP, and relative LnCPI = Log of domestic consumer price
interest rates on exchange rates in the long term index; LnNER = Log of nominal exchange
and short term. There is an insignificant positive rate; LnUSWPI = Log of import price index;
relationship between oil prices and the exchange LnOILP = Log of international crude oil prices;
rate in the long term and a significant positive LnRGDPG = Log of real output growth; 𝛽0 =
relationship in the short term. There is an Constant term or the intercept of the model;
insignificant negative effect between the money 𝛽1 - 𝛽4 = Elasticities or the expected effect
supply relative to the exchange rate in the long of their respective independent variables; εt =
term and a significant negative relationship in the Error term
short term.
The coefficients 𝛽1 represent the impact of
3. DATA AND METHODOLOGY exchange rate on the Consumer Price Index
(CPI), specifically capturing the influence of
3.1 Model Specification exchange rate changes on the prices of imported
goods and domestically produced tradable goods,
In line with other studies such as Ari et al. (2021); which can be affected by variations in the prices
Abiodun et al. [16], we consider controlling for of imported inputs.

30
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

3.2 Linear ARDL Model Nigeria. During the study period, the average
exchange rate in Nigeria was N210.9095. The
Equation (3) is a static model which does not minimum was N93.41000 and the maximum
capture the economic reality of the dynamic recorded was N6 96.0100, with the standard
impact exchange rate might have on domestic deviation indicating high level of exchange rate
price level. To capture this dynamic impact of volatility in the country. The average import price
exchange rate on domestic price level, we index was $127.7632, while the minimum and
employ the autoregressive distributed lag (ARDL) the maximum was $109.6000 and $165.9700
model of Pesaran, Shin and Smith [22]. The respectively, indicating high cost of importation in
above equation (3) is now specified in ARDL Nigeria during the study period. The average
form as: crude oil price was $62.50344 while the minimum
and the maximum crude oil prices were 14.28000
∆Ln(CPI)t = 𝛼0 + ∑ni=1 𝛼1i ∆Ln(CPI)t-i+ ∑ni=0𝛼2i and $130.1000, respectively. The average real
∆Ln(NER)t-i + ∑ni=0 𝛼3i ∆Ln(USWPI)t-i+ ∑ni=0 𝛼4i GDP growth during the study period was
∆Ln(OILP)t-i+ ∑ni=0 𝛼5i ∆Ln(RGDPG)t-i + 𝛽1 1.359112 percent, while the maximum and the
Ln(CPI)t-i + 𝛽2 Ln(NER)t-i+ 𝛽3 Ln(USWPI)t-i+ minimum were 3.600000 percent and -1.800000
𝛽4 Ln(OILP)t-i+𝛽6 Ln(RGDPG)t-i + εt 4 percent, respectively, implying that the real GDP
growth was not constant throughout the
3.3 Technique of Estimation observed timeframe in Nigeria.

Prior to deciding on the appropriate econometric The residuals of CPI are normally distributed,
estimation technique, the study tests the whereas the residuals of all other variables have
individual variables for unit root using Augmented high rate of dispersion from their mean (with P-
Dickey-Fuller (ADF) unit root test. Consequent to values less than 0.05 at the 5% significant level)
this test, cointegration test using ARDL Bound which is evidenced in the Jarque Bera statistics.
testing approach was employed. Then the study The results also suggest that all the variables are
estimates both the short-run and the long-run positively skewed except real GDP growth, which
versions of the ARDL model. The study implies that the data is not distributed
employed ARDL model. Autoregressive symmetrically.
distributed-lag modelis a model which captures
the dynamic relationship between the dependent 4.2 Correlation Analysis
and the independent variables over time. The
choice of this model is based on the fact that it The correlation analysis is employed to ascertain
does not require all variables to be integrated of the strength of the linear relationship among the
same order. Thus, variables can be stationary at explanatory variables of the model. The essence
level and first difference. of this is to ensure that multicollinearity does not
constitute a serious problem. These results are
4. RESULTS AND DISCUSSION summarized in Table 2.

4.1 Descriptive Statistics The result reveals that all the variables are
positively correlated. However, it was found that
Table 1 presents the summary of descriptive no pair-wise correlation coefficient between any
statistics of the dependent variable and the two regressors was up to the stipulated
explanatory variables included in the model. benchmark of 0.8 except for USWPI and
NER. However, this will not invalidate our
Results in Table 1 shows that the average entire regression results based on our specific
inflation rate (CPI) during the study period was research context and the goals of our analysis.
14.20927, while the minimum and the maximum Therefore, we conclude that no serious
level recorded during this period was 8.160000 problem of multicollinearity exists in the
and 21.47000 respectively. The minimum level sample.
shows that the lowest level of inflation
experienced in the country during the period of 4.3 Data Analyses: Pre-Estimation Test
the study was 8.160000 percent while the
highest level was 21.47000 percent, which is The analyses start with the test for unit root. This
almost two-fold of the average indicated above, is followed by the estimation of the long run
thus showing the level of price instability in relationship among the variables.

31
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

Table 1. Summary of descriptive statistics

CPI (%) NER USWPI OILP RGDPG

Mean 14.20927 210.9095 127.7632 62.50344 1.359112
Median 14.28000 181.8050 124.4000 61.53500 2.065593
Maximum 21.47000 696.0100 165.9700 130.1000 3.600000
Minimum 8.160000 93.41000 109.6000 14.28000 -1.800000
Std. Dev. 3.458795 125.4330 15.01392 20.50544 1.943643
Skewness 0.123709 2.194614 1.224027 0.966521 -0.666986
Kurtosis 2.011297 7.183233 3.542012 4.608964 1.931522
Jarque-Bera 4.154999 147.0590 25.14697 25.30166 11.68451
Probability 0.125243 0.000000 0.000003 0.000003 0.002902
Sum 1364.090 20247.31 12265.27 6000.330 130.4748
Sum Sq. Dev. 1136.510 1494676. 21414.68 39944.92 358.8860
Observations 96 96 96 96 96
Source: Researcher’s Computation (2023) using E-views 10

Table 2. Summary of correlation analysis

Correlation Probability CPI NER USWPI LnOILP RGDPG

CPI I.000000
----
NER 0.624920 1.000000
(0.0000) ----
USWPI 0.579071 0.902761 1.000000
(0.0000) (0.0000) ----
LnOILP 0.273089 0.531739 0.622982 1.000000
(00071) (0.0000) (0.0000) ----
RGDPG 0.028462 0.347595 0.437640 0.684385 1.000000
(0.7831) (0.0005) (0.0000) (0.0000) (0.0000)
Source: Researcher’s Computation (2023), using E-Views 10

4.3.1 Unit root test this study used the ARDL Bounds test. The
hypothesis of the Bounds test states that;
The Table 3 below is a summary of the unit root
test on the data at level and first difference. The H0: There is no long run relationship existing
augmented dickey fuller (ADF) unit root test amongst the variables.
was performed to ascertain the order of
integration. Thus, the critical values and H1: There is long run relationship amongst
probability values are shown in open brackets the variables.
and parentheses respectively below the ADF test
statistics. The test result is summarized below:

Results of the unit root test using Augmented Under the null hypothesis of no long-run
Dickey Fuller procedure shows that all the relationship, the result indicates that the null
variables are stationary at first difference with an hypothesis cannot be accepted. This is because
exception of oil price which is stationary at level. the computed F-statistic of about 4.21 is
This supports the use of autoregressive greater than the upper bounds I(1) critical value
distribution lag (ARDL) model. However, the of 3.49 at the 5% significance level. This implies
presence of a unit root in the series that all the variables are co-integrated. Based
suggests the need to carry out cointegration on this, we conclude that a long-run
test. relationship exists between domestic price level
and its fundamentals.
4.3.2 ARDL bounds test (F-test for
Cointegration) 4.4 Model Estimation and Evaluation

In order to determine the existence or otherwise Given that we have confirmed the existence of
of a long-run relationship among the variables, long run relationship amongst the variables, we

32
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

then estimate the short-run and long-run 4.4.2 Evaluation of estimated linear short-run
parameters of the Linear ARDL model. results

4.4.1 Evaluation of estimated linear long-run The short-term relationship between the
results exchange rate and consumer price inflation in
this study was estimated using the appropriate
The estimated long-run coefficients of our linear lag order. The distributed lag (DL) of the
model is reported in Table 5. autoregressive (AR) model determined that 4 lag
periods were selected for the dependent variable
Table 5 shows that on average, a percent (consumer price inflation), 3 lag periods for the
positive change in nominal exchange rate (NER) nominal exchange rate. 2 lag periods were
is associated with about 0.024 percent increase chosen for the import price index (USWPI), 1 lag
in Nigeria’s domestic prices in the long run. The period for crude oil prices and 4 lag periods for
p value of 0.03 shows that the positive effect is real output growth. However, the aim of error
significant in the long run at 5% level. This correction modeling (ECM) is to assess whether
finding corroborate the findings of Musa [9] on a short-term disequilibrium can be corrected in
exchange rate vis-à-vis consumer price index. the long run. In this context, the error correction
term serves as an indicator of the speed at which
On the average, import prices also has a positive adjustments occur from one period to another. It
coefficient of about 2.9666, which shows that is expected to exhibit a negative sign, fall within
import prices in Nigeria exert positive influence the range of 0 and -1, and be statistically
on domestic price level in the long run. significant at the 5% level to demonstrate a
Consequent upon this, we can assert that in the robust convergence process towards long-run
long-run, a unit percent change in import prices equilibrium. The detailed estimation results can
increase domestic prices on the average by be found in Table 6.
about 2.9666 percent. The p-value of 0.8078
indicates that the positive effect of import price The estimated short-term coefficients in Table 6
index on domestic price inflation in the long run is indicate a speed of adjustment of 0.069762 units
not significant at 5% level. The findings of Usman per month. This means that the model corrects
and Mohammad [12] is supported by the findings any deviations from equilibrium from the previous
of this study. month at a rate of approximately 0.0698 units
each month. Essentially, if the exchange rate
The long run effect of oil price on domestic price remains constant at this monthly rate, the
level in Nigeria is negative. This is evident in the domestic price level will gradually reach its long-
negative coefficient of -1.231525. What this run equilibrium. This adjustment occurs when
entails is that a unit percent increase in oil price there are changes in the exchange rate, such as
per barrel, on average, results to 0.4688 percent an appreciation or depreciation of the domestic
decrease in domestic prices in the long run. The currency. The adjustment process takes place
decrease was also found to be insignificant at over several months, with approximately 0.0698
5% level. This is in tandem with the findings of units of adjustment occurring each month. The
Tela et al. [13] which established that crude oil error correction term, which is correctly signed
prices is positively related to domestic price level. and statistically significant, indicates that in
Nigeria, any deviations from equilibrium in the
Real output growth on the other hand decreases short run will be corrected in the opposite
domestic prices in the long run but not direction over time.
significantly in Nigeria. This implies there is
0.769113 percent decrease on the average in We can also see from the result that a unit
domestic prices, when real output growth in the percent increase in domestic inflation (CPI)
Nigeria economy increases by a unit percent. continues to have a significant influence on its
The negative relationship between real output own growth in Nigeria. The positive and
growth and domestic prices is in contrast with the significant lagged coefficients indicate that a unit
findings of Abiodun et al. [16]. percent increase in previous one month's
inflation on an average, contributes
The intercept in the long run estimate is positive approximately 0.3902 percent to its current
(0.257155). This implies that there is a positive month’s growth. On average, increases of
trend in domestic price inflation in Nigeria around 0.1888 percent in CPI can be attributed
occasioned by factors, other than the variables to a unit percent increase in inflation rate
included in the model. observed two months ago, while an average

33
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

increase of approximately 0.3111 percent can be [11], but contradicts the findings of Musa [9];
attributed to a unit percent increase in inflation Usman and Mohammad (2016); and Victor
rate observed three months ago. (2016), who found positive significant impact of
exchange rate on domestic prices in Nigeria.
The observed data suggests also, that a unit
percent positive change in the nominal exchange A unit percent increase in the import price index
rate has a small negative but statistically on average, triggers a rise in the domestic price
significant impact on inflation (CPI). Specifically, level in Nigeria by approximately 5.8411 percent
on an average, for every unit percent increase in in the current month. However, the three
the nominal exchange rate, there is a decrease preceding months’ unit percent positive change
in CPI by approximately 0.0005 percent in the on average, leads positive trends in Nigeria’s
current month. Furthermore, the data shows that domestic inflation, with the first, second, and third
this impact is not limited to the current month but previous months lags increasing the domestic
extends to the preceding two months as well, price level (CPI) by approximately 4.8273,
where it accounts for decreases of 0.0003 and 7.0781, and 6.7169 percent respectively. The
0.0002 percent in CPI, respectively. However, probability values associated with the import
the statistical analysis revealed a probability price index for the current month, second lagged
value of 0.6405. This value suggests that the month, and third lagged month indicate that their
negative trend observed in the nominal exchange positive short-term impact on the current month
rate is not statistically significant at the 5% level inflation in Nigeria is statistically significant.
in the current month. In contrast, the negative However, the positive short-term effect in the first
trend was found to have shown significance in lagged month was not statistically significant at
the previous two months.Meanwhile, the finding the 5% level. This finding is in tandem with the
corroborates the result from Gbalam and Dumani findings of Usman and Mohammad (2016).

Table 3. Summary of ADF unit root test results

Variables At Level Level of At First Diff Order of Decision

Significance Integration
CPI -1.5988 5% -3.9138** I(1) Stationary
(-2.8936) (-2.8936)
{0.4791} {0.0029}
NER 2.7903 5% -4.3821** I(1) Stationary
(-2.8929) (-2.8932)
{1.0000} {0.0006}
USWPI 2.2791 5% -8.3784** I(1) Stationary
(-2.8922) (-2.8925)
{1.0000} {0.0000}
OILP -3.0315 5% -7.6852** (- I(0) Stationary
(-2.8925) 2.8929)
{0.0356} {0.0000}
RGDPG -2.1632 5% -9.5920** (- I(1) Stationary
(-2.8922) 2.8925)
{0.2211} {0.0000}
Source: Researcher’s Compilation (2023), using E-Views 10
Note: {} = Probability values
() = Critical values
** =Significant at 5%, Obtained from MacKinnon (1996)

Table 4. Summary of linear ARDL bounds test results

F-Statistic Significance Critical Values Conclusion

Lower Bound Upper Bound
4.210932 10% 2.2 3.09 Cointegrated at
5% 2.56 3.49 10%, 5% &
2.5% 2.88 3.87 2.5% levels
1% 3.29 4.37
Source: Researcher’s Computation (2023), using E-Views 10

34
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

The short-run coefficient of oil price is positive and recommendation purposes. Specifically, we
although it is not significant at 5% level. This tested for serial correlation, Heteroscedasticity,
means, on the average, a unit percent increase normality and model stability.
in international crude oil prices will cause inflation
to increase by about 0.2655 percent in Nigeria. 4.5.1 Serial correlation test
This corroborates the findings of Tela et al. [13]
which established that crude oil price is positively We test for the presence of serial correlation
related to domestic price level. among the explanatory variables in the model
using the Breusch-Godfery LM test. The test is
The coefficient (0.025556) of real output growth conducted under the following hypotheses.
in the current month leads to a rise in domestic
price level on average, by that percentage for a H0: Residuals are not serially correlated in
unit percent increase in real output growth in the ARDL model (no autocorrelation)
Nigeria, whereas the coefficient from the
previous month triggers about 0.0547 percent H1: Residuals are serially correlated in the
decrease in Nigeria’s consumer price inflation ARDL Model (autocorrelation)
(CPI) for a unit percent growth in real outputs.
The probability values of 0.4845 and 0.1654 From Table 7, the probability of the F-statistic in
indicate that the observed positive and negative the models is greater than 0.05 (5% significant
effects of real output growth on CPI in Nigeria, level). This implies the null hypothesis of no
were not statistically significant at the 5% autocorrelation or no serial correlation in the
level.The negative relationship between real ARDL model is accepted at the 5 percent level of
output growth and domestic price level supports significance, hence we conclude there is no
the findings of Abiodun et al. [16]. autocorrelation in the model. However, the
results of the LM test for serial correlation further
Overall, a lack of significance does not corroborate the Durbin Watson test-statistic
necessarily mean that there is no relationship result for serial correlation.
between the variables; it simply indicates that the
observed relationship (given the available data 4.5.2 Heteroscedasticity test
used for the analysis) is not statistically strong
enough to establish a confident conclusion. In Table 8, we present the results of the
heteroscedasticity test for the model using the
The coefficient of determination (R2) is reported Breusch-Pagan LM approach, in order to see
as 0.7425. This means on the average, whether the error variance of each observation is
approximately 74 percent of the changes in the fixed or not. The null hypothesis of no
domestic price level in Nigeria can be attributed heteroscedasticity against its alternative
to the combined influence of the nominal hypothesis is:
exchange rate, import prices, international crude
oil prices, and real output growth. The remaining H0: Error term of the ARDL model is
26 percent of the variation in consumer price homoscedastic (there is no
inflation is likely due to other factors not captured heteroscedasticity in the residuals)
in the model (residuals). These findings indicate
that the model has a relatively high explanatory H1: Error term of the ARDL model is not
power and a strong ability to predict, suggesting homoscedastic (there is heteroscedasticity in
that it fits well with the available data. From the the residuals)
regression estimates, it can be further observed
that the calculated F-statistic is 4.210932. This We reject the null hypothesis if the Probability
shows that the variables are jointly statistically Chi-Square of the Observed*R-Squared is less
significant at 5 percent.The Durbin Watson (DW) than 0.05 (5% significant level). The results of
Statistic in the model is approximately equal to 2, the model indicate that we cannot reject the null
suggesting that there is no autocorrelation hypothesis, since the Probability of Chi-Square of
problem in the model. the Observed*R-square is greater than 0.05. This
suggests that the estimated model does not have
4.5 Post Estimation Diagnostic Tests the issue of heteroskedasticity at the 5 per cent
level of significance, which is a good signal that
Relevant post-estimation diagnostic tests were the ARDL model is homoscedastic and adequate
conducted to check whether the estimated model over the period covered by this study and
is robust, suitable and valid for policy application therefore, the data is reliable for predication.

35
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

Table 5. Summary of estimated linear long-run results

Variable Coefficient Std. Error t-Statistic Prob.

NER 0.024087 0.010919 2.205974 0.0305
LnUSWPI 2.966561 12.15318 0.244097 0.8078
LnOILP -1.231525 2.519424 -0.488812 0.6264
RGDPG -0.769113 0.398485 -1.930094 0.0575
C 0.257155 55.80244 0.004608 0.9963
Note: not all variables are in log, (** = significant at the 5% level)
Source: Researcher’s Computation (2023), using E-Views 10

Table 6. Summary of estimated linear short-run result

Variable Coefficient Std. Error t-Statistic Prob.

D(CPI(-1)) 0.390174** 0.099816 3.908914 0.0002
D(CPI(-2)) 0.188752 0.107367 1.758001 0.0829
D(CPI(-3)) 0.311056** 0.099126 3.137969 0.0025
D(NER) -0.000514 0.001096 -0.468927 0.6405
D(NER(-1)) -0.003224** 0.001077 -2.993095 0.0038
D(NER(-2)) -0.002817** 0.001171 -2.404392 0.0187
D(LnUSWPI) 5.841110** 2.720192 2.147315 0.0351
D(LnUSWPI(-1)) 4.827323 2.744554 1.758873 0.0828
D(LnUSWPI(-2)) 7.078102** 2.772270 2.553179 0.0128
D(LnUSWPI(-3)) 6.716908** 2.717758 2.471489 0.0158
D(LnOILP) 0.265530 0.196875 1.348719 0.1816
D(RGDPG) 0.025556 0.036367 0.702726 0.4845
D(RGDPG(-1)) -0.054712 0.039044 -1.401294 0.1654
CointEq(-1)* -0.069762** 0.013427 -5.195778 0.0000
R-squared 0.742549
Adjusted R-squared 0.699640
Durbin-Watson stat 1.780019
Note: not all variables are in log, (** = significant at the 5%)
Source: Researcher’s Computation (2023), using E-Views 10

Table 7. Results of serial correlation test

F-statistic 2.510442 Prob. F (2,71) 0.0884

Obs*R-squared 6.076243 Prob. Chi-Square (2) 0.0479
Source: Researcher’s Computation (2023), using E-Views 10

Table 8. Results of heteroscedasticity test

F-statistic 0.519488 Prob. F (18,73) 0.9403

Obs*R-squared 10.44643 Prob. Chi-Square (18) 0.9164
Scaled explained SS 7.982608 Prob. Chi-Square (18) 0.9789
Source: Researcher’s Computation (2023), using E-Views 10

5. CONCLUSION AND POLICY as a result of increases in import prices which

RECOMMENDATIONS probably had origin in endogenous demand
shock and exogenous foreign producers’ pricing
Given the institutional framework in Nigeria which shocks, which cause the marginal costs of
is characterized by reduced productivity, policy Nigerian businessmen and producers to
inconsistencies and macro-economic increase. As a result of existing inflation
fluctuations, the rise in domestic prices has persistence in the country as observed from the
become part and parcel of the Nigerian short-run analysis, businessmen in Nigeria are
economy. However, based on the findings in able to pass the increases in marginal costs of
(4.8), inflation in Nigeria during the recent time is production on to Nigeria consumers in the form

36
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

of increases in the prices of consumer durables 7. Dornbusch R, Fischer S. Macroeconomics

and non-durables in the country. This study (6th ed.). McGraw-Hill; 1993.
therefore concludes that in recent time, 8. Mordi CNO, Adeleke AI. An empirical
exchange rate has insignificant positive long-run analysis of the effect of exchange rate
relationship with domestic price level in Nigeria, fluctuations on inflation in Nigeria.
domestic prices responds swiftly to import prices European Journal of Business and
reflecting Nigeria’s trading partners price levels, Management. 2014;6(17):171-181.
and that, positive and negative changes in the 9. Musa N. Impact of exchange rate volatility
nominal exchange rate have similar effects on on inflation in Nigeria. Journal of
domestic prices during the period studied; Contemporary Research in Business,
January 2015 to December 2022. Economics and Finance. 2021;3(1):26-38.
10. Ari A, Edson M, Félix FS. An empirical
5.1 Policy Recommendations assessment of the exchange rate pass-
through in Mozambique. IMF Working
i. Governments at all levels in the country Paper, © 2021International Monetary Fund,
should encourage and support the W.P. 2021;21-132I.
innovative ideas of business firms and 11. Gbalam PE, Dumani M. Foreign exchange
individuals. This will support local rate and consumer price changes in the
production, hence, there can be Nigerian economy. Saudi Journal of
substitution of imported goods for Economics and Finance. 2020;4(2):64-71.
domestically produced goods and the 12. Usman O, Muhammad SM. Revisiting
exchange rate will be stable. exchange rate pass-through to consumer
ii. The monetary authority’s recent efforts to price inflation in Nigeria: A cointegrated
move towards greater flexibility in the vector autoregressive approach. Academic
exchange rate should be part of a Journal of Economic Studies. 2018;
consistent and credible package of 4(1):60–67.
sustainable policies, since the impact of 13. Tela SA, Adesoye AB, Ologundudu MM.
exchange rate on domestic price level in The exchange rate fluctuations, domestic
Nigeria is insignificant. prices and monetary policy in Nigeria.
International Journal of Economics,
COMPETING INTERESTS Commerce and Management United
Kingdom. 2018;6(10).
Authors have declared that no competing 14. Helmy O, Fayed M, Hussien K. Exchange
interests exist. rate pass-through to inflation in Egypt: A
structural VAR approach. Review of
REFERENCES Economics and Political Science.
2018;3(2):2-19.
1. Mankiw NG. Principles of economics. 15. Hernán R, Norberto R. Nonlinear state and
Cengage learning; 2016. shock dependence of exchange rate pass-
2. Mankiw NG, Taylor MP. Principles of through on Prices. Monetary and
macroeconomics. Cengage Learning; 2017. Economic Department, BIS Working.
3. Bawa S, Abdullahi IS. Threshold effect of 2018;690.
inflation on economic growth in Nigeria. 16. Abiodun SB, Ajibola IO, Inuwa AT, Idowu P,
CBN Journal of Applied Statistics. 2012; Sani B, Anigwe JO, Udoko CO. Exchange
3(1):43–63. rate pass-through to inflation in Nigeria.
4. National bureau of statistics; 2021. CBN Journal of Applied Statistics. 2016;
Available:www.nigerianstat.gov.ng 7(1a):49-70.
5. Ijeoma OP. Impact of exchange rate 17. Victor OA. The causes of persistent
variation and inflation on economic growth inflation in Nigeria. CBN Journal of Applied
in Nigeria: An empirical approach. Statistics. 2016;7(2).
Research Journal of Finance and 18. Lariau A, El Said M, Takebe M. An
Accounting. 2014;5(22). assessment of the exchange rate pass-
6. Nwankwo O, Okafor R. Exchange rate through in Angola and Nigeria. IMF
pass-through to domestic prices in Nigeria: Working Paper. 2016;01-191.
A disaggregated analysis. Journal of 19. Cheikh NB, Zaied YB, Ameur HB, Recent
Economics and Sustainable Development. developments in exchange rate pass-
2013;4(1):21-34. through: What have we learned from

37
 Okechukwu et al.; Asian Res. J. Arts Soc. Sci., vol. 21, no. 3, pp. 26-38, 2023; Article no.ARJASS.107165

uncertain times?. Journal of International Price And Exchange Rate In Indonesia.

Money and Finance. 2023;131:102805 International Journal of Energy Economics
20. Hassan TA, Mertens TM, Zhang TA, Risk- and Policy. 2023;13(2):27.
based theory of exchange rate stabilization. 22. Pesaran M, Shin Y, Smith R.
The Review of Economic Studies. Bounds testing approaches to
2023;90(2):879-911 the analysis of level relationships. Journal
21. Andriyanto RW, Tresnaningtyas A, of Applied Econometrics. 2001;17:
Sumanda K, Palupi WA. Analysis of Oil 289-326.

© 2023 Okechukwu et al.; This is an Open Access article distributed under the terms of the Creative Commons Attribution
License (http://creativecommons.org/licenses/by/4.0), which permits unrestricted use, distribution, and reproduction in any
medium, provided the original work is properly cited.

Peer-review history:
The peer review history for this paper can be accessed here:
https://www.sdiarticle5.com/review-history/107165

38

View publication stats