Exchange Rate Pass-Through on Prices in Nigeria—A Threshold Analysis

Abstract

Persistent exchange rate depreciation and its debilitating effects on rising inflation have remained a concern in Nigeria. This article explores the effects of exchange rate pass-through on producer prices, consumer prices, export prices, import prices and the Taylor rule from 2000 to 2023, using quarterly data and adopting threshold autoregression and self-exciting smooth threshold regression methods. The findings suggest that there are non-linearities in the way that exchange rate depreciation affects prices in Nigeria. Furthermore, the findings suggest a threshold of 5 percent depreciation. Two sub-sample analyses corroborate the main findings, showing that a threshold of 5 percent is the optimum benchmark if demand and supply are not to be weakened. At this level or below, the effects of exchange rate depreciation on inflation are much lower, even though prices will rise. However, above this benchmark, the effects of depreciation on inflation are much larger, weakening consumer demand for both imported and domestic goods as well as producer supply of both exported and domestic goods and services in the economy. This result implies that an average exchange rate depreciation not higher than 5 percent within a quarter is reasonable if the Nigerian economy is to remain competitive both domestically and globally. Finally, the results suggest that the exchange rate pass-through to prices is considerably higher in Nigeria below the threshold, while it overshoots for producer prices, export prices, and import prices above the threshold. To keep inflation in check, this paper suggests that the monetary authorities should try to keep exchange rate depreciation below the established thresholds, while also considering adjusting the policy rate to take into account the exchange rate depreciation thresholds in order to keep domestic prices stable.

1. Introduction

Over the years, exchange rate management has become a central topic of discussion in monetary economics. This is because the currency exchange rate framework serves as the medium through which a country connects to the international market, by facilitating the trade in goods and services between two or more countries. Also, the exchange rate transmission channel is the link through which external shocks affect an economy and represents the transmission medium through which domestic shocks are mostly transmitted to other economies. Taking a cue from the Russian war in Ukraine, the ensuing rise in commodity prices in food and non-food items, such as wheat, crude oil, and many others, are a result of the geopolitical tensions in Eastern Europe, leading to rising import prices for countries that depend most on these items (especially an African country like Nigeria). Furthermore, COVID-19 is another typical example of how the global transmission of goods and services can affect the exchange rate mechanism of monetary policy. Nigeria, as an import-dependent economy, has struggled to stabilize its exchange rate policies over the years.

In the year 2000, the Nigerian Naira was pegged at roughly USD 1/NGN 100 (CBN 2023). Due to the constant demand for foreign currency and the dearth of supply to meet this demand, the country switched to a managed float exchange rate system in 2014. Between 2014 and May 2023, Nigeria adopted the managed float exchange rate system, allowing the central bank to keep the exchange rate value at a value they determined to be the fair value of the Naira in the official market, which was mostly overvalued compared to the unofficial market exchange rate (parallel market or black market). As of the end of 2022, Nigeria’s exchange rate traded at USD 1/NGN 461.5 and USD 1/NGN 735 in the official and unofficial markets (CBN 2023), respectively, depicting the overvaluation of the Naira in the official market and the different valuations of the exchange rate across different foreign exchange markets. More recently, after the new government came to power in May 2023, it implemented a uniform exchange rate system in June 2023. This system enables individuals who meet the criteria to access foreign exchange through the investors and exporters window (I&E FX window) at a price that is determined by market demand and supply. Among other benefits, the exchange rate unification reform of 2023 was designed to consolidate all exchange rates into a single market-determined exchange rate, thereby enhancing transparency and price discovery in the foreign exchange market. As a result of this, the exchange rate depreciated to USD 1/NGN 899.89 in December 2023, before descending further to USD 1/NGN 1601 as of the end of September 2024 (CBN 2024), further highlighting the exchange rate instability issues the country currently faces. This has led to a rise in import prices and worsening domestic inflation, as inflation numbers rose to decade highs of 34.19 percent in June 2024 before easing slightly to 33.40 percent in July 2024.

How an economy manages its exchange rate framework is an important criterion for how it can achieve price stability, particularly in a developing economy such as Nigeria, which is import-dependent for almost every consumer and producer item. By implication, external events tend to affect domestic prices of goods and services in import-dependent economies, leading to exchange rate vulnerability and price fluctuations, thereby affecting macroeconomic performance. To manage price fluctuations, the Central Bank of Nigeria (CBN) announced in 2009 that it would implement an inflation-targeting strategy to effectively control inflation and ensure that inflation remains at single digits while maintaining sustainable economic growth (CBN 2021). After 14 years, the country has still not provided a transition framework that will help it achieve its primary macroeconomic objectives, such as reducing inflation, boosting economic growth, and creating more jobs. Rather, the inflation-targeting activities have been implied in the CBN’s communique and its activities. Perhaps the fact that a framework has not yet been established has been a reason why the CBN has been unable to steer the economy in the right direction, as prices continue to rise despite its regular interventions. Moreover, the macroeconomic situation has become more difficult over this 14-year hiatus, with Nigeria experiencing two recessions in the last decade and rapidly rising inflation to decade highs of over 34 percent as of the time of writing.

Inflation targeting is a kind of monetary policy where the authorities publicly declare and commit to a medium-term goal for the inflation rate. The theory behind this approach to monetary policy is that for monetary policy to be successful, it must foster long-term growth while preserving price stability via inflation management. Interest rates are the primary tool the CBN employs to influence the economy in the short term. The importance of the inflation-targeting strategy has been acknowledged in many previous studies (Abango et al. 2019; Nasir et al. 2020; Oyadeyi 2024a; Valogo et al. 2023). However, this has not been well implemented in Nigeria, evidenced by the worsening economic conditions. Furthermore, this does not pertain to Nigeria alone, as many other developing countries continually face the challenge of inflation caused by currency depreciation, even though the inflation-targeting strategy accommodates the role of the exchange rate within the monetary policy rule to account for shocks from external events that may influence price stability as a result of swings in the exchange rate (Alagidede and Ibrahim 2017; Castellares and Toma 2020; Bhat et al. 2022; Oyadeyi 2022a; Valogo et al. 2023).

Consequently, questions arise regarding the importance of a monetary policy rule based on the currency depreciation threshold that is necessary within the policy rule for the efficacy of implementing monetary policy. It, therefore, becomes imperative for us to determine the optimum thresholds beyond which depreciation in the Nigerian currency worsens domestic prices and impoverishes households, consumers, and businesses. To buttress the point, Figure 1 shows the yearly inflation target set by the CBN between 2000 and 2023, and how it performs against these targets. As the graph reveals, the monetary authorities were only able to meet their average inflation targets in the years 2004, 2006, 2007, 2010, 2012, 2013, 2014, and 2018, while they have not been able to achieve these targets since 2019, and this has been a major cause for concern in recent years. This implies that they met their target eight times in 24 years. Moreover, the Nigerian economy has not been able to meet the developing economies’ target since 2016. A nuanced perspective also shows that this target was only achieved once since the last two recessions (Nigeria had its first recession in the last decade in 2016 due to global falling crude oil prices. Its second recession occurred during the COVID-19 pandemic in 2020). The period since the two recessions has been marred by rising inflation rates post-2018, as demonstrated in Figure 1. Therefore, achieving inflation targets remains a central issue in the discourse of monetary policy in Nigeria today.

Several studies have observed the exchange rate pass-through on prices globally (Nasir and Vo 2020; Rigobon 2020; and many others), while many others have focused on developing countries (De Mendonça and Tiberto 2017; Kassi et al. 2019; Cheikh and Zaied 2020). However, some have focused on country-specific studies (Anghelescu 2022; Hong et al. 2022; Yigit 2022). What is common among the results of these studies is that the pass-through seems to be generally low and incomplete in the short run, and it tends to be higher in the long term. This implies that, to some degree, the rate at which a country exchanges its currency has a defining role in determining the price level of commodities in an economy. In Nigeria, several studies have embarked on this discussion. Many of these previous studies have focused on using the linear and non-linear implications of exchange rate pass-through on consumer prices and import prices (Adekunle et al. 2019; Adedokun et al. 2022) within the non-linear autoregressive framework, while others have focused on the transmission of the pass-through (Ogundipe and Egbetokun 2013; Bada et al. 2016; Bello and Sanusi 2019); these studies found a considerable effect of exchange rate pass-through on consumer prices in Nigeria.

Based on the sample of the literature reviewed, this paper identified that none of the previous empirical studies on Nigeria has examined the exchange rate pass-through effect on producer prices. This effect is important because the effect of the exchange rate on producer prices has a signalling effect on productive capacity and consumer demand in any economy. Therefore, an empirical examination of this effect may help us understand to what extent the pass-through affects both producer and consumer prices in Nigeria and their degree of elasticity. Furthermore, this paper also identified that previous studies have only focused on import prices, forgetting that the prices of exports are important to the Nigerian economy since exports have historically been the main driver of government revenue. Even though the prices of exports have not been a central concern in other studies, it is very important to establish how currency depreciation affects our exports. This is because it may give us an indication of what is to happen if Nigeria begins to diversify its exports as suggested in the 2022–2024 Medium Term Expenditure Framework. Perhaps, the pass-through may also give us an indication of why Nigeria’s exports (apart from crude oil) have not been globally competitive.

Finally, any discussion on the threshold effects beyond which an exchange rate depreciation hurts prices is lacking for Nigeria. While previous studies have investigated the effects of currency depreciation on consumer and import prices, the threshold level beyond which currency depreciation would hurt prices significantly has not yet been empirically investigated for Nigeria. The importance of this is because, ceteris paribus, an exchange rate depreciation will increase the prices of imported items and lower the prices of exported items, thereby raising the cost of producing items and reducing the cost of exporting these products on the one hand, while increasing the costs of imported goods and raising the prices of goods and services domestically, thereby reducing the purchasing power of consumers on the other hand. Moreover, the purchasing power of consumers and business enterprises that rely substantially on imported raw materials, semi-finished goods, and finished goods is negatively affected by the fluctuating value of the Naira. Therefore, the optimum threshold value beyond which a currency’s depreciation becomes unfavourable for consumer demand and producer supply is yet to be investigated by previous studies on the subject matter, calling for further research. In addition, recent studies in line with the updated Taylor-type rule suggest that, given the increased economic integration of modern economies, central banks should incorporate exchange rate performance into their reaction functions when determining interest rate changes (Valogo et al. 2023). To that end, the study also analyzes the optimal level of currency depreciation that should be factored into the CBN monetary policy framework in Nigeria.

Therefore, the following objectives are central to the discussion in this paper. First, the paper will explore the optimal threshold level beyond which a depreciation of the exchange rate starts to hurt producer and consumer prices in Nigeria. This exchange rate depreciation resulting in a rise in prices can create financial pressure on households, especially the poor and vulnerable. Second, the paper will establish the optimum threshold level beyond which exchange rate depreciation harms export and import prices in Nigeria. It is important to understand the difference between these two objectives. On the one hand, it can be beneficial for export and import prices to increase after a currency depreciation (so long as it is within the threshold), since this may help correct trade imbalances by making imports more expensive and unattractive, thereby lowering demand for imports, and raising exports, as they become cheaper globally when the exchange rate depreciates, thereby making them more attractive. On the other hand, consumer and producer prices should not rise beyond a certain threshold. This is because when prices go up, the costs of production and consumption also increase, creating financial pressures on businesses and households, as businesses may want to pass these costs on to the final consumers, creating a chain reaction as households reduce their demand for these items, which also affects the businesses that produce these items. Consequently, establishing an exchange rate threshold for prices is vital to maintaining price stability in the economy. Hence, monetary policy attention is required to understand this distinction, especially when formulating monetary policies. This is why this study will undertake this research regarding the optimum level beyond which a depreciation of the Naira will hurt prices in Nigeria in achieving price stability. Finally, the paper will analyse how currency depreciation affects the Central Bank of Nigeria’s monetary policy decisions and responses (reaction functions).

The findings of this paper contribute significantly to the literature by addressing this underexplored area of exchange rate pass-through on export and producer prices in Nigeria through the critical examination of their optimum threshold levels using two distinct threshold models, thereby resolving a critical gap in the existing research. Nigeria, an emerging market with significant exchange rate volatility, offers a valuable case study that improves our comprehension of the influence of currency fluctuations on various pricing channels. This research expands upon the current emphasis on advanced economies by providing a view of the operation of exchange rate pass-through in less stable economies, which experience more severe currency instability and inflationary pressures. The paper not only informs monetary and fiscal policies for managing inflation and stabilizing exports by analysing Nigeria’s distinctive economic environment but also provides a framework that can be applied to other emergent markets with comparable characteristics. These findings contribute to the advancement of the discourse on exchange rate economics and provide valuable lessons for both academicians and policymakers worldwide by enriching the comprehension of how exchange rate pass-through operates in economies with diverse financial and trade structures in a broader global context. The rest of the paper is designed as follows: the literature review is discussed in the second section, the estimation techniques are discussed in the third section, the following section discusses the results, and the final section concludes the paper.

2. Literature Review

In advanced countries, Kiliç (2016) found that there is the existence of a smooth transition type of pass-through on prices, depending on the regime, in the economies of Japan, Canada, the United Kingdom (UK), Germany, Australia, and the United States of America (USA). The paper adopted a logistic smooth transition regression to achieve the objectives. The paper suggested that the pass-through process is complete in the long run, while it is incomplete over the short term. Furthermore, the study found that the pass-through process was found to affect prices asymmetrically. Bandt and Razafindrabe (2014) and many other researchers have discovered the same findings: an imperfect pass-through of exchange rates on prices in the short run but a full pass-through on prices in the long run across nations in the Euro area. In addition, the study argued that the global financial crisis had a role in accelerating the upward trend in exchange rate pass-through on prices in these nations. Auer and Schoenle (2016) focused on firm-level pricing behaviour in reaction to the pass-through in the US. Their paper found that firms that change their prices more often when there is a change in their input costs do not change their prices when the prices of their import competitors change. The paper also found a U-shaped curve for import prices and a bell-shaped curve for the prices of their import competitors, in line with the propositions of Dornbusch (1987).

Further, regarding advanced countries, Jiménez-Rodrígueza and Morales-Zumaquero (2016) focused on the G-7 countries. The findings of their study corroborated earlier studies that signalled that the pass-through simply depends on the regime, while there was a significant connection within the nexus among these countries. Furthermore, Choudhri et al. (2005) explored the exchange rate pass-through within the G-7 countries (except the US). The paper used hybrid models, as well as individual models such as producer and local currency pricing, to test this relationship. Their findings suggested that the outcomes from the hybrid model are stronger than the other models for G-7 countries such as France, the UK, Germany, Japan, Canada, and Italy. Nasir and Vo (2020) examined this link with respect to the first three countries that adopted the inflation-targeting strategy (New Zealand, the UK, and Canada). Contrary to previous findings, which stated that the currency pass-through has led to a reduction in inflation in these countries, the paper suggested that these countries’ pass-through to inflation had increased as a result of adopting the inflation targeting strategy. For Switzerland, Oktay (2022) found a low pass-through in the Swiss economy. On the other hand, Ali and Anwar (2016) showed that exchange rate depreciation can help resolve the price puzzle.

For panel data sets combining both developed and developing nations, Devereux et al. (2004) suggested that money supply volatility is an important characteristic of the outcomes of currency pass-through in an economy. Therefore, they suggested that countries with high money volatility growth would witness high levels of pass-through, and vice versa. Also, Takhtamanova (2010) found that an inelastic pass-through was a result of the lower levels of inflation in the 1990s among 14 selected OECD countries. Using state-space models, López-Villavicencio and Pourroy (2019) found an imperfect pass-through for countries currently adopting the inflation-targeting strategy, while inflation-targeting effectiveness tended to be higher when the inflation targets were higher and complemented by lower levels of monetary independence. This outcome of a lower pass-through was in line with another study by Ozkan and Erden (2015), which focused on a combination of 88 advanced and developing nations.

As for panel studies focusing entirely on developing markets, De Mendonça and Tiberto (2017), in a study of 118 developing countries, posited that monetary policy transparency is essential to mediate the negative role of currency depreciation on inflation. In a sample of 14 emerging market countries, López-Villavicencio and Mignon (2017) found that the adoption of an inflation-targeting strategy as well as transparent monetary policy operations lowered the pass-through on inflation. Devereux and Yetman (2010) found that countries with low inflation generally have a low pass-through. The paper further highlighted that the major reason for this is the slow process of adjusting prices within these countries. Cheikh and Zaied (2020) examined the currency depreciation on prices for ten countries that had recently joined the EU. Their paper also found that a commitment to inflation targeting lowers the level of inflation among these countries, while it found a threshold of 4.56 percent, below which the pass-through is reduced and above which it leads to a complete adjustment in high-inflation countries.

Using a panel threshold on 24 developing countries, Cheikh and Louhichi (2016) suggested that the adjustment was higher in countries with high inflation, and vice versa. In a panel of Asian countries, Kassi et al. (2019) found that the adjustment is relatively higher in emerging Asian countries whose inflation is relatively lower and less volatile than in developing Asian countries. Aleem and Lahiani (2014) found a lower pass-through in East Asia and Latin America. Also, the study suggested that the pass-through was a bit higher in Latin America compared to East Asia.

Among studies that focused on country-specific emerging markets and developing countries, Nasir et al. (2020) examined the currency depreciation effect in the Czech Republic. The study showed that exchange rates have important roles to play in inflation expectations and the current inflation in the Czech Republic. On the other hand, Casas (2020) focused on manufacturing firms and found heterogeneous impacts of currency depreciation on manufacturing firms’ prices in Colombia. Also, Anghelescu (2022) found heterogeneous exchange rate impacts in Romania, while monetary policy shocks caused the largest changes in exchange rates for Romania. In Vietnam, however, Hong et al. (2022) suggested that money supply changes, as well as the rate at which the currency is exchanged, play a prominent role in inducing prices in Vietnam. Furthermore, they found asymmetries in the way consumer prices respond to currency changes.

Furthermore, among the country-specific studies of emerging markets and developing countries, Yigit (2022) found that exchange rate shocks and domestic inflation shocks were key determinants of inflation in Turkey. Khundrakpam (2008) suggested a higher adjustment of prices in India until recent years prior to 2008, due to the recent changes in policies and management frameworks. On the other hand, Bhat et al. (2022) also investigated the pass-through in India. The paper found that, in the immediate term, an exchange rate depreciation reduces prices when economic activities are generally low, while appreciation increases prices when the Indian economy is booming. However, the study found the pass-through to be generally incomplete for India. In addition, Castellares and Toma (2020) examined the effects of exchange rate adjustment on the Peruvian economy. The paper disaggregated consumer prices into their different categories and investigated the effect of currency rates and the pricing law on the disaggregated prices. They found that for non-durable items with dollarized pricing, exchange rate pass-through had no effect whatsoever, while it had partial effects on durable items with dollarized prices.

As for country-specific studies that focus on African countries, Kabundi and Mlachila (2019) found that, in South Africa, monetary policy independence and credibility had lowered the impact of a depreciation in the South African currency on domestic prices during the period when the country implemented the inflation-targeting framework. Furthermore, Fandamu et al. (2023) found a non-linear and incomplete adjustment in Zambia using the structural vector autoregressive (SVAR) framework. This implies that depreciation or appreciation will affect consumer prices differently for Zambia. Valogo et al. (2023) is one of the few empirical studies that has established the threshold effect of a currency depreciation on domestic prices within a specific developing country. Focusing on Ghana, the paper found that at a threshold of 0.70 percent or above, the exchange rate positively impacts prices in Ghana. Furthermore, the study found that despite a threshold of 0.51 percent, exchange rates have positive effects on the monetary policy of Ghana.

Finally, this paper reviews past studies on the subject matter that focus on Nigeria due to our interest in the country. In Nigeria, studies like Adedokun et al. (2022) found, using the linear and non-linear framework, that the exchange rate is a better predictor of inflation than the money supply. Studies such as Ogundipe and Egbetokun (2013) corroborate this finding. Furthermore, Adekunle et al. (2019) found that currency depreciation had a heterogeneous effect on inflation and imported inflation, and that it diminishes the adjustment process. Also, the study found a partial and imperfect nexus within the short term, while it became larger in the long term. This result also corroborates an earlier finding by Bada et al. (2016), who suggested an incomplete adjustment in import and consumer prices for Nigeria. Oyinlola and Babatunde (2009) were of the view that global export prices had a larger effect on Nigeria’s inflation compared to the exchange rates. On the other hand, Bello and Sanusi (2019) suggested that firms’ marginal cost, energy inflation, food inflation, and imported inflation are key determinants of Nigeria’s inflation. However, Adelowokan (2012) did not find any evidence of exchange rate pass-through on inflation in Nigeria.

In summary, the literature review demonstrates that the exchange rate pass-through on prices has been well researched, both domestically and internationally. However, the nuances of the pass-through on export prices and producer prices in Nigeria needs more empirical investigation. Furthermore, based on the findings of the literature review showing a lack of studies focusing on Nigeria specifically, this study will include a more focused discussion on the unique structural characteristics of Nigeria, such as its heavy import dependency and dual exchange rate regime, and how these differ from those of other countries in terms of their impact on exchange rate pass-through. Exchange rate fluctuations have a potentially more significant impact on inflation in Nigeria than in countries with more diversified domestic production due to the country’s dependence on imported products, particularly on essentials such as petroleum and food. The pass-through effect is further exacerbated by the dual exchange rate system, which frequently relies on the parallel market rate, resulting in a more volatile price-setting behaviour. In addition, the study will compare the dynamics of Nigeria’s exchange rate pass-through with those of other emergent markets that may have more robust monetary policies or inflation-targeting frameworks. This will provide a more nuanced understanding of how exchange rate pass-through operates under various structural conditions, as well as position Nigeria’s experience within the broader global context. Therefore, this study will undertake an investigation to ascertain the threshold effects of exchange rate pass-through on producer prices, import prices, consumer prices, and export prices to provide a more nuanced argument on how exchange rates affect prices in Nigeria.

3. Materials and Methods

3.1. Theoretical Framework

The theoretical framework of the study rests on the monetary policy perspective on exchange rate pass-through and how it affects inflation in an economy. The extent to which exchange rate pass-through affects domestic prices is significantly influenced by monetary policy. The exchange rate pass-through of countries with credible inflation-targeting regimes is frequently lower, as the need for firms to modify prices in response to exchange rate fluctuations is reduced by anchored inflation expectations. Central banks with robust price stability mandates can further mitigate exchange rate pass-through by employing instruments, such as interest rate adjustments, to regulate inflation, reduce demand, and stabilise the currency. Exchange rate pass-through is also influenced by the type of exchange rate regime in operation, with floating regimes typically enduring a higher pass-through than fixed or managed float systems. Exchange rate pass-through is typically lower in countries with low inflation because price stability mitigates the effects of fluctuations in exchange rates on domestic prices. In contrast, it may be elevated in situations of elevated inflation and less credible monetary policies. Additionally, there may be an asymmetry in pass-through, where the central bank’s response to inflationary pressures may result in a stronger impact on pricing from exchange rate depreciation than from appreciation. Monetary policy’s efficacy in stabilising inflation and expectations is crucial for managing exchange rate pass-through in an economy.

3.2. Empirical Model Specification

Using the monetary policy perspective on exchange rate pass-through, this paper builds a model within the Taylor rule framework to achieve its objectives. We follow the works of Valogo et al. (2023), Bernanke et al. (2018), and Aleem and Lahiani (2014) in accounting for the central bank’s role in modelling the pass-through:

$$pr{i}_t^{*}={\tau }_0+{\tau }_{1}ex{c}_t+{\tau }_{2}mp{r}_t+{\epsilon }_t$$

(1)

where pri* represents a vector of variables meant to capture prices in Nigeria (consumer prices, export prices, producer prices, and import prices), exc represents exchange rate, and mpr represents the monetary policy rate, which is the main tool the central bank uses to adjust prices in Nigeria (Oyadeyi 2022b, 2024b). ε_t is the normally distributed error term, while $\tau$’s are the parameters to be estimated. Including the monetary policy rate within the exchange rate model is meant to capture the influence of central banks in trying to control the inflation rate. Moreover, monetary policy operations are central to understanding the relationship between currency depreciation and inflation in any economy. Following Aleem and Lahiani (2014), we augment Equation (1) to include output and the lags of the regressors within our model. The inclusion of these is meant to establish how past information about the regressors influences the inflation rate, since policymakers depend on this while formulating monetary policies. Therefore, Equation (1) would be specified as follows:

$$pr{i}_t^{*}={\tau }_0+{\tau }_{1}ex{c}_t+{\tau }_{2}ex{c}_{t-1}+{\tau }_{3}mp{r}_t+{\tau }_{4}mp{r}_{t-1}+{\tau }_5{y}_t+{\tau }_6{y}_{t-1}+{\epsilon }_t$$

(2)

where $\tau$’s are the parameters of concern, and mpr_t−1, y_t−1, and exc_t−1 are the lags of the regressors, represented as the lags of the monetary policy rate, real GDP growth, and exchange rate, respectively. pri* represents a vector of variables meant to capture prices. In our case, prices are represented by producer prices, consumer prices, export prices, and import prices.

Following the works of Valogo et al. (2023), the paper modifies the Taylor rule propounded by Taylor (1993) to account for the output and inflation gap within the monetary policy framework as follows:

$$mp{r}_t={\tau }_0+{\tau }_{1}mp{r}_{t-1}+{\tau }_{2}ex{c}_t+{\tau }_{3}ex{c}_{t-1}+{\tau }_4(in{f}_t-in{f}_t^{*})+{\tau }_5(y_t-y_t^{*})+{\epsilon }_t$$

(3)

where (inf_t − inf_t*) is the inflation gap, (y_t − y_t*) represents the output gap, and exc represents the exchange rate, while mpr represents the monetary policy rate. The lags of mpr and exc were included to account for expectations in policy decisions based on past information about the variables. τ₂ and τ₃ are the weights attached when the monetary authorities want to make decisions on inflation by accounting for exchange rate movements within their model. The reason why the monetary policy rate and inflation gaps are considered within this model is that to maintain inflation and output, central banks frequently alter interest rates in response to fluctuations in exchange rates. The Taylor rule encapsulates the method by which a central bank systematically modifies its policy interest rate in response to deviations from the target inflation rate and output gap. Consequently, the inclusion of the monetary policy rate in the model contributes to the understanding of how monetary policy actions either amplify or mitigate the direct impact of exchange rate changes on domestic prices.

Additionally, inflation expectations can be influenced by exchange rate fluctuations, particularly in open economies. The Taylor rule’s emphasis on inflation enables a direct connection between inflation targeting and exchange rate pass-through, which is indicative of the level of well-anchored inflation expectations and the effectiveness of monetary policy in mitigating inflationary pressures related to exchange rate fluctuations. The inflation gap, which is the discrepancy between actual and target inflation, is of paramount importance, as it dictates the degree to which a central bank may react to inflationary pressures resulting from fluctuations in exchange rates. If the inflation differential is substantial, the central bank may raise interest rates, which would have a direct impact on domestic prices and would mitigate the pass-through effect. In essence, monetary policy actions may not be exogenous; they frequently respond to inflationary pressures that result from fluctuations in exchange rates. Therefore, by incorporating monetary policy and the inflation gap into the analysis, this endogeneity is mitigated, thereby guaranteeing that the measured exchange rate pass-through accurately represents the actual impact of exchange rate fluctuations on prices, rather than the secondary effects of monetary policy adjustments. From Equation (3), a direct relationship with mpr would imply that the exchange rate, its lag, and the lag of the inflation rate would have the same sign. If this has been established, then it stands to reason that a policy variable might be used to target both the inflation rate and the movement in exchange rate (Ozdemir 2020; Valogo et al. 2023).

3.3. Estimation Technique

This paper uses two variants of the threshold technique. These include the threshold autoregressive technique (TAR) model proposed by Tong (1978) and the smooth threshold regression (STR) to estimate the objectives. The TAR model effectively captures non-linear, regime-switching behaviour in time series data by allowing the dynamics of the series to vary when a threshold is crossed. It encapsulates the asymmetric dynamics of a process that typical linear models cannot accommodate. On the other hand, the STR model is an econometric technique that is employed to identify non-linear relationships between variables. In this model, the relationship’s evolution is contingent upon a continuous transition between regimes, rather than a sudden change. It is especially beneficial for the analysis of situations in which the behaviour of a dependent variable undergoes a progressive change as an independent variable surpasses a specific threshold. This paper follows the works of Valogo et al. (2023) in adopting this technique to confirm the existence of asymmetry and generate the optimum currency depreciation threshold for the exchange rate to become an important criterion in setting monetary policy decisions. Furthermore, the method splits the sample into at least two regimes using an exogenous variable—in our case, the exchange rate—where the two regimes represent asymmetry in the data findings. The reason for adopting the TAR model is that it assumes the transition to the alternative regimes may be abrupt and that we may need to fix the threshold manually, which may sometimes be subject to bias. STR, on the other hand, assumes the transition to alternative regimes is smooth, either exponential or logistic. Therefore, these two methods were adopted to examine the threshold effects as stated in the objectives. Setting up the model in line with Munir et al. (2009), therefore, the threshold specification can be specified as follows:

$${\lambda }_t=({\mathsf{\ell }}_{01}+{\displaystyle \sum _{i-1}^p{\mathsf{\ell }}_{1i}{K}_t})d[ex{c}_t\le T]+\left({\mathsf{\ell }}_{02}+{\displaystyle \sum _{i=1}^p{\mathsf{\ell }}_{2i}{K}_t}\right)d[ex{c}_t>T]+{\epsilon }_t$$

(4)

where ${\lambda }_t$ represents the dependent variable. In our case, ${\lambda }_t$ represents a vector of variables meant to capture prices in Equation (2) and the monetary policy rate in Equation (3), while K_t represents a vector of independent variables, including their lags. T represents the threshold indicator, which splits the results into a lower and upper regime. Due to the unique interpretations of the model, a depreciation above the threshold value would imply a high regime, while a depreciation below this implies a low regime. $ex{c}_t>T$ and $ex{c}_t\le T$ represent the high and low regimes (asymmetries), whereas ${\mathsf{\ell }}_{01},{\mathsf{\ell }}_{02},{\mathsf{\ell }}_{1i},$ and ${\mathsf{\ell }}_{2i}$ are the regime-dependent parameters to be analysed. Therefore, Equation (4) can be re-specified as follows:

$$\begin{array}{l}pr{i}_t^{*}=\left({\tau }_{01}+{\tau }_{11}ex{c}_t+{\tau }_{12}ex{c}_{t-1}+{\tau }_{13}mp{r}_t+{\tau }_{14}mp{r}_{t-1}+{\tau }_{15}{y}_t+{\tau }_{16}{y}_{t-1}\right)d\left[ex{c}_t\le T\right]+\\ \left({\tau }_{02}+{\tau }_{21}ex{c}_t+{\tau }_{22}ex{c}_{t-1}+{\tau }_{23}mp{r}_t+{\tau }_{24}mp{r}_{t-1}+{\tau }_{25}{y}_t+{\tau }_{26}{y}_{t-1}\right)d\left[ex{c}_t>T\right]+{\epsilon }_t\end{array}$$

(5)

$$\begin{array}{l}mp{r}_t=\left[{\iota }_{01}+{\iota }_{11}mp{r}_{t-1}+{\iota }_{12}\left({\inf }_t-{\inf }_t^{*}\right)+{\iota }_{13}\left(y_t-y_t^{*}\right)+{\iota }_{14}ex{c}_t+{\iota }_{15}ex{c}_{t-1}\right]d\langle ex{c}_t\le T\rangle +\\ \left[{\iota }_{11}+{\iota }_{21}mp{r}_{t-1}+{\iota }_{22}\left({\inf }_t-{\inf }_t^{*}\right)+{\iota }_{23}\left(y_t-y_t^{*}\right)+{\iota }_{24}ex{c}_t+{\iota }_{25}ex{c}_{t-1}\right]d\langle ex{c}_t>T\rangle +{\epsilon }_t\end{array}$$

(6)

where all variables are as previously described. The Wald (1943) test was used to test the non-linearity of our models. The reason for using the Wald test to test the non-linearity of the models is because it is a critical component of threshold regression analysis, since it evaluates the significance of the threshold effect or the prevalence of nonlinearity in the relationship between the dependent and independent variables. In particular, it evaluates whether the model with the threshold effect (i.e., regime switching) considerably enhances the model’s fit in comparison to a linear model without the threshold. Therefore, the Wald test is specified with a null hypothesis of equality between the low and high regimes, while the alternative states otherwise. For the robustness of the estimates, the paper used the self-exciting smooth threshold regression to validate the threshold results. To better understand and capture the steps in the use of the TAR and STR models, the summary below describes their usage, particularly as it relates to exchange rate pass-through.

The TAR and STR models are highly effective instruments for the analysis of exchange rate pass-through to domestic prices, as well as for the identification of nonlinear relationships and regime shifts in time series data. The model operates by approximating a criterion variable (such as the exchange rate in our case) and establishing a threshold value at which the data are divided into distinct regimes. The relationship between exchange rates and prices is reflected in the distinct behaviours of these regimes. For example, exchange rate pass-through may be elevated during periods of elevated inflation or economic instability, but it may be diminished in more stable environments. These models accurately depict the dynamics between domestic prices and exchange rates by capturing these abrupt shifts between regimes. In addition, these models enable researchers to identify nonlinear pass-through effects in the context of exchange rate pass-through, distinguishing periods of high and low pass-through based on economic conditions.

For instance, the influence of exchange rate fluctuations on prices may be more pronounced during periods of elevated inflation, whereas it may be less pronounced during periods of low inflation or stable economic conditions. Therefore, these models are essential for policymakers to effectively manage inflation and stabilise prices, as they provide insights into the extent to which various factors, such as exchange rate fluctuations and monetary policy decisions, influence pass-through by estimating regime-specific coefficients. Therefore, the steps to take in using the TAR and STR methods include the following. First, identify the threshold variable (exchange rate), the non-threshold variables (other variables which affect this relationship, provided in Equations (5) and (6)), and the dependent variable (which are consumer prices, producer prices, and import and export prices in this paper). The second step includes estimating the model using the TAR and STR methods to find the threshold value. A value below the threshold represents the lower regime, while any results above this threshold represent the upper regime. The final step includes the interpretation of the results.

3.4. Data Sources and Description

Table 1. Variable descriptions, sources, and measurements.

Variables and Abbreviation	Measurements	Expected Sign	Sources	References
Consumer Price (LCPI)	The log of consumer price index	-	CBN Bulletin 2023	(Ozdemir 2020; Valogo et al. 2023; Aleem and Lahiani 2014)
Producer Price (LPPI)	The log of GDP Deflator	-	CBN Bulletin 2023	(Jiménez-Rodrígueza and Morales-Zumaquero 2016; Casas 2020)
Export Price (LEPI)	The log of export price index	-	CBN Bulletin 2023	(Jiménez-Rodrígueza and Morales-Zumaquero 2016; Casas 2020)
Import Price (LIPI)	The log of import price index	-	CBN Bulletin 2023	(Ozdemir 2020; Valogo et al. 2023; Aleem and Lahiani 2014)
Exchange Rate (LEXC)	The log of nominal exchange rate	Positive on prices and the policy rate	CBN Bulletin 2023	(Ozdemir 2020; Valogo et al. 2023; Aleem and Lahiani 2014)
Monetary Policy (MPR)	The benchmark interest rate is the percentage that central banks charge commercial banks as a fee for overnight loans of surplus cash from their reserve accounts.	Negative on prices	CBN Bulletin 2023	(Ozdemir 2020; Valogo et al. 2023; Aleem and Lahiani 2014)
Inflation Gap (INFGAP)	The difference between actual inflation rate and the trended inflation rate (inf − inf*).	Positive on the policy rate	CBN Bulletin 2023	(Ozdemir 2020; Valogo et al. 2023; Aleem and Lahiani 2014)
Output Gap (YGAP)	The difference between actual real GDP growth and the trended real GDP growth (y − y*)	Positive on the policy rate	CBN Bulletin 2023	(Ozdemir 2020; Valogo et al. 2023; Aleem and Lahiani 2014)
Output (RGDPG)	The real GDP Growth	Negative on prices	CBN Bulletin 2023	(Ozdemir 2020; Valogo et al. 2023)

depicts the data that were used in achieving the study’s objectives. The data were sourced from the CBN statistical bulletin, 2023 edition, spanning the years 2000 to 2023 on a quarterly basis. The output and inflation gaps were derived using the Hodrick-Prescott Filter, while the other variables were used as collated from the bulletin.

4. Empirical Results and Discussion

4.1. Preliminary Analysis

This paper starts its analysis by taking a descriptive characterisation of the data set. These results are presented in

Table 2. Summary statistics of the data.

	LCPI	LPPI	LEPI	LIPI	MPR	INFGAP	YGAP	RGDPG	LEXC
Mean	4.80	6.25	4.77	4.83	12.73	0.00	0.00	4.50	5.20
Median	4.80	5.48	4.69	4.71	13.00	−0.40	0.14	4.74	5.04
Maximum	6.21	8.27	5.64	5.63	20.50	11.34	12.41	19.17	6.29
Minimum	3.40	4.61	4.23	4.29	6.00	−12.60	−14.12	−7.59	4.61
Std. Dev.	0.76	1.39	0.29	0.27	3.25	3.79	2.96	3.82	0.45
Skewness	0.03	0.25	0.91	1.01	0.05	0.02	−0.44	0.12	0.76
Kurtosis	1.98	1.25	3.28	3.82	3.33	4.05	10.88	5.84	2.18
Jarque-Bera	4.01	12.76	12.86	18.37	0.47	4.25	241.05	31.05	11.42
Probability	0.13	0.18	0.89	0.56	0.79	0.12	0.40	0.43	0.31
Sum	441.4	574.6	438.7	444.0	1171.3	0.00	0.00	413.7	478.1
Sum Sq. Dev.	52.2	176.3	7.5	6.6	963.8	1310.1	799.8	1326.7	18.2
Observations	96	96	96	96	96	96	96	96	96

Source: author’s computations. Note: LCPI is the log of the consumer price index, LPPI is the log of the GDP Deflator, LEPI is the log of the export price index, LIPI is the log of the import price index, MPR is the monetary policy rate, INFGAP is the inflation gap, YGAP is the real GDP growth gap, and LEXC is the log of the nominal exchange rate.

. Table 2 shows that there are 96 observations in each of the data sets, while each of the observed variables was within its lower and upper bound values. The average real GDP growth was 4.5 percent, while its median growth was 4.74 percent. The results further showed that the log of consumer prices grew at an average of 4.80 percent per quarter, while the log of producer prices grew by 6.23 percent on average per quarter. On the other hand, the log of export prices grew by 4.77 percent per quarter, while the log of import prices grew by 4.83 percent per quarter. The study showed that the log of exchange rate depreciation averaged 5.20 percent per quarter, while the monetary policy rate averaged 12.73 percent.

The skewness statistic revealed that except for the output gap, the rest of the variables were positively skewed. Positive skewness in a dataset suggests that the tail of the distribution is elongated to the right. This implies that the majority of the data values are concentrated on the left side of the distribution, with a lengthy tail that extends to the right. That is, the mean is typically greater than the median in a positively skewed distribution, as the mean is more influenced by the larger values in the tail. The kurtosis of LEPI, LIPI, MPR, INFGAP, YGAP, and RGDPG exceeded 3, meaning that they follow a leptokurtic distribution. This means that these variables are highly peaked in relation to the normal or mesokurtic distribution. On the other hand, LCPI, LPPI, and LEXC were less than 3, meaning they follow a platykurtic distribution. This means that they are less peaked in relation to the normal distribution. Finally, the Jarque-Bera statistics showed that the variables assume normality with their datasets.

Table 3. Variance inflation factor.

Consumer Prices			Producer Prices
Panel A	Centred		Panel B	Centred
Variable	VIF	1/VIF	Variable	VIF	1/VIF
C	NA		C	NA
LEXC	4.2277	0.2365	LEXC	4.2277	0.2365
LEXC_1	4.1644	0.2401	LEXC_1	4.1644	0.2401
MPR	9.3494	0.1070	MPR	9.3494	0.1070
MPR_1	9.1451	0.1093	MPR_1	9.1451	0.1093
RGDPG	1.0881	0.9191	RGDPG	1.0881	0.9191
RGDPG_1	1.1353	0.8808	RGDPG_1	1.1353	0.8808
Export Prices			Import Prices
Panel C	Centred		Panel D	Centred
Variable	VIF	1/VIF	Variable	VIF	1/VIF
C	NA		C	NA
LEXC	4.2277	0.2365	LEXC	4.2277	0.2365
LEXC_1	4.1644	0.2401	LEXC_1	4.1644	0.2401
MPR	3.3494	0.1070	MPR	3.3494	0.1070
MPR_1	3.1451	0.1093	MPR_1	3.1451	0.1093
RGDPG	1.0881	0.9191	RGDPG	1.0881	0.9191
RGDPG_1	1.1353	0.8808	RGDPG_1	1.1353	0.8808
Taylor Rule
Panel E	Centred
Variable	VIF	1/VIF
C	NA
LEXC	4.3638	0.2292
INFGAP	1.0725	0.9324
YGAP	1.0757	0.9296
MPR_1	1.2734	0.7853
LEXC_1	4.0402	0.2475

Source: author’s computation.

presents the variable inflation factor meant to capture the extent of multicollinearity within the regression models. A tolerance factor of more than 10 means that multicollinearity among the dataset is high, while some scholars put this figure at 5. Therefore, any dataset that has a multicollinearity of values greater than 5 was removed from the regression. Using either the 5 or 10 benchmark, the centred variance inflation factor was below 5, showing that none of the variables exhibited the problems of multicollinearity. This is because the tolerance values of the key variables range from 0.1 to 0.92 in Panels A to D, while they range from 0.23 to 0.93 in Panel E of Table 3. Finally,

Table 4. Augmented Dickey-Fuller and Phillip-Perron unit root tests.

Variables	ADF	PP	Remark
	I(0)	I(0)
LCPI	−3.7164 ***	−10.3617 ***	Stationary
LPPI	−9.3344 ***	−9.3344 ***	Stationary
LEPI	−11.5473 ***	−18.3423 ***	Stationary
LIPI	−10.1019 ***	−11.4374 ***	Stationary
LEXC	−5.6079 ***	−5.1313 ***	Stationary
MPR	−8.1210 ***	−8.1636 ***	Stationary
INFGAP	−8.3237 ***	−11.5558 ***	Stationary
YGAP	−9.2723 ***	−19.0772 ***	Stationary
RGDPG	−9.1062 ***	−16.3350 ***	Stationary

Source: author’s computations. Note: the t-statistics of the unit root tests were reported, and they are significant at −3.50 for 1 percent, where ‘***’ indicate significance at 1 percent.

presents the unit root characteristics as a preliminary estimation for the main results. This paper employed the Augmented Dickey-Fuller and Phillip-Perron unit root tests to establish if the variables had a unit root. The outcomes show that the variables are all I(0), meaning they were stationary in their level form. A major requirement to employ the threshold regression method is for the variables to be stationary in their level form. Given the absence of non-stationary variables, the paper proceeds with the threshold analysis.

4.2. Main Analysis

4.2.1. The Wald Test

In starting its main analysis, this paper employed the Wald test to confirm if non-linearity exists within the specified models. With the probability values of 0.0000 shown in

Table 5. Wald tests.

Model	Chi-Square Statistic	p-Value
Exchange Rate Pass-through to consumer prices	368.8556	0.0000 ***
Exchange Rate Pass-through to producer prices	1123.55	0.0000 ***
Exchange Rate Pass-through to export prices	9712.544	0.0000 ***
Exchange Rate Pass-through to import prices	7954.658	0.0000 ***
Taylor Rule	272.5593	0.0000 ***

Source: author’s computation. Note: ‘***’ indicate significance at 1 percent.

, this implies that non-linearity exists within the models. Therefore, this paper rejects the null hypothesis, which states that the models are linear in the low and high regimes. This implies that higher currency depreciation tends to affect prices more than periods of lower currency depreciation. This may be because of the macroeconomic fundamentals driving business decisions and confidence in the Nigerian economy.

4.2.2. Exchange Rate Pass-Through on Producer and Consumer Prices

In employing the threshold regression model, this paper employed the TAR and STR models in achieving the objectives. The essence of the smooth threshold regression is to capture the non-linear impacts of the threshold results and to confirm if the results are in line with the discreet threshold results, which separates the results into two regimes. The lower and upper regimes in the discreet threshold regression also represent non-linearity, as opined by Valogo et al. (2023). These results are reported in

Table 6. The threshold results of exchange rate pass-through on consumer and producer prices.

	Exchange Rate Pass-Through on Consumer Prices				Exchange Rate Pass-Through on Producer Prices
	Threshold Regression		Smooth Threshold Regression		Threshold Regression		Smooth Threshold Regression
	Low Regime	High Regime	Linear Model	Non-Linear	Low Regime	High Regime	Linear Model	Non-Linear
LEXC	0.7343	0.8404	0.7286	0.8402	0.7302	1.4928	0.7308	1.5289
	(0.0047) ***	(0.0013) ***	(0.0060) ***	(0.0000) ***	(0.0033) ***	(0.0049) ***	(0.0065) ***	(0.0000) ***
	Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors
LEXC(-1)	0.0355 **	(0.0172)	0.0271 ***	(0.0129)	0.1647 *	(0.0844)	0.2145 *	(0.0802)
MPR	−0.0356 ***	(0.0012)	−0.0391 ***	(0.0004)	−0.0879 **	(0.0121)	−0.0869 **	(0.0166)
MPR(-1)	−0.0227 **	(0.0382)	−0.0204 *	(0.0651)	−0.0345 **	(0.0336)	−0.0348 **	(0.0353)
RGDPG	0.0057 ***	(0.0055)	0.0058 ***	(0.0015)	0.0026 ***	(0.0084)	0.0076 ***	(0.0054)
RGDPG(-1)	0.0114 ***	(0.0030)	0.0086 **	(0.0190)	0.0031 ***	(0.0080)	0.0091 ***	(0.0043)
C	−0.4779 *	(0.0923)			2.1363 **	(0.0133)
Threshold	5.008		5.0046 ***	(0.0000)	5.0095		5.0088 ***	(0.0000)
SSR	1.0804		1.0377		11.3751		11.2473
Regime	2		2		2		2
Serial Correlation	0.2214		0.2547		0.6058		0.229
Heteroscedasticity	0.1694		0.2281		0.1968		0.235
Normality Test	0.8913		0.484		0.8003		0.1897

Source: author’s computations. Note: The probability values are in parenthesis while the coefficients are not in parenthesis. ‘***’, ‘**’, and ‘*’ are significance at 1 percent, 5 percent and 10 percent, respectively.

. As for the impact on consumer prices, the findings established a threshold of 5 percent for a quarterly exchange rate depreciation for both the TAR and STR models. This implies that annual exchange rate depreciation must not exceed 20 percent; otherwise, it will lead to more debilitating inflation. Given that Nigeria has seen the average annual exchange rate depreciate at roughly this same amount each quarter in the past and before the recent administration began its tenure in 2023, this looks like a reasonable threshold value. Interestingly, this paper found that the effect on consumer prices had a positive and significant effect in both regimes of the TAR and STR models.

From both the TAR and STR models, the results report a significant and positive effect of currency depreciation on consumer prices, meaning that when an exchange rate depreciation occurs, it increases prices, on average, by 73 percent and 84 percent in the low and upper regimes of the TAR, respectively. Alternatively, it does so by 73 percent and 84 percent in the linear and non-linear models of the STR, respectively. These findings may be due to the dependent nature of the Nigerian economy. Therefore, a depreciation results in a weakening of the domestic currency against the international benchmark, leading to a rise in import prices and thereby raising consumer prices. Monetary policy, both in the previous and current periods, has had significant negative effects of currency depreciation on prices, while output has a positive effect in both the TAR and STR models. These results imply that the effect of currency depreciation on consumer prices is imperfect, asymmetric, and to a higher degree in Nigeria.

The results of the threshold for currency depreciation on producer prices were identical to those of consumer prices. This implies that at a threshold of 5 percent, exchange rate depreciation would have a positive effect on producer prices in both the TAR model and the STR model. Regarding output, when exchange rate depreciation is lower than the threshold, it empirically reduces inflation as expected in the TAR and STR models. On the other hand, the current monetary policy reduces prices as expected, since monetary policy is the main policy tool used by the central bank to control inflation. Finally, the results of the post-diagnostic tests using the Breusch-Godfrey serial correlation test, Jarque-Bera normality tests, and the heteroscedasticity tests using the Breusch-Pagan-Godfrey tests showed that the models are free from serial correlation, are normally distributed, and are not heteroscedastic.

4.2.3. Exchange Rate Pass-Through on Export and Import Prices

The second part of the main analysis focuses on the effect of currency depreciation on export and import prices. The results are reported in

Table 7. The threshold results of exchange rate pass-through on export and import prices.

	Exchange Rate Pass-Through on Export Prices				Exchange Rate Pass-Through on Import Prices
	Threshold Regression		Smooth Threshold Regression		Threshold Regression		Smooth Threshold Regression
	Low Regime	High Regime	Linear Model	Non-Linear	Low Regime	High Regime	Linear Model	Non-Linear
LEXC	0.6319 ***	1.5722 ***	0.6319 ***	1.6059 ***	0.7716 ***	1.6500 **	0.7696 **	1.6490 ***
	(0.0057)	(0.0085)	(0.0001)	(0.0000)	(0.0091)	(0.0308)	(0.0426)	(0.0000)
	Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors
LEXC(-1)	0.2720 **	(0.0471)	0.8665 *	(0.0642)	0.1440 *	(0.0665)	0.2381 *	(0.0518)
MPR	−0.0075 **	(0.0481)	−0.0223 **	(0.0205)	−0.0180 **	(0.0135)	−0.0239 *	(0.0734)
MPR(-1)	−0.0137 **	(0.0324)	−0.0446 **	(0.0176)	−0.0181 **	(0.0147)	−0.0268 **	(0.0474)
RGDPG	0.0060 **	(0.0223)	0.0167 **	(0.0127)	0.0016 *	(0.0719)	0.0004 *	(0.0926)
RGDPG(-1)	0.0017 **	(0.0286)	0.0017 *	(0.0797)	0.0041 **	(0.0347)	0.0028 *	(0.0562)
C	3.1965 ***	(0.0000)	3.1523 ***	(0.0000)	1.6324 ***	(0.0058)	1.6459 ***	(0.0051)
Threshold	5.0072		5.0152 ***	(0.0000)	5.0119		5.0214 ***	(0.0000)
SSR	1.6853		2.85		1.3506		3.28
Regime	2		2		2		2
Serial Correlation	0.1657		0.277		0.138		0.1267
Heteroscedasticity	0.334		0.6018		0.614		0.5476
Normality Test	0.812		0.7657		0.3681		0.6456

Source: author’s computations. Note: The probability values are in parenthesis while the coefficients are not in parenthesis. ‘***’, ‘**’, and ‘*’ are significance at 1 percent, 5 percent and 10 percent, respectively.

. The TAR and STR results found a threshold of roughly 5 percent for a quarterly currency depreciation on export and import prices. This implies that a maximum depreciation of roughly 20 percent within a given year may not negatively impact consumer purchasing power and businesses in the economy. Interestingly, our analysis found that exchange rate adjustments on export and import prices had a positive and significant effect on the TAR and STR models. This implies that exchange rates below the threshold have a milder effect on rising prices, unlike exchange rate depreciation above the threshold, which tends to have a more debilitating impact on prices in the economy.

Indeed, exchange rates within the low and high regimes affect prices differently according to the STR model. However, when this depreciation is within the threshold of 5 percent, it has a milder effect on prices in the economy, unlike when it is above the threshold, which has a stronger and more harmful effect on prices and economic activity. The lag of monetary policy had significant negative effects on export prices in the STR and TAR models, as expected, while output had significant positive effects on prices within the threshold. Finally, the results of the post-diagnostic tests using the Breusch-Godfrey serial correlation test, Jarque-Bera normality tests, and the heteroscedasticity tests using the Breusch-Pagan-Godfrey tests showed that the models are free from serial correlation, are normally distributed, and are not heteroscedastic.

4.2.4. Exchange Rate Depreciation within the Taylor Rule of Monetary Policy

As illustrated in

Table 8. Exchange rate thresholds in the Taylor rule results.

	Threshold Regression		Smooth Threshold Regression
	Low Regime	High Regime	Linear Model	Non-Linear
LEXC	3.5474 **	5.8211 **	1.6465 ***	5.0711 ***
	(0.0122)	(0.0149)	(0.0022)	(0.0019)
	Non-Threshold Regressors		Non-Threshold Regressors
INFGAP	0.0111 *	(0.0729)	0.2080 **	(0.0207)
YGAP	0.0250 *	(0.0539)	0.0868 **	(0.0171)
MPR(-1)	−0.9010 ***	(0.0000)	−0.2835 **	(0.0272)
LEXC(-1)	3.0036 **	(0.0232)	1.1104 **	(0.0133)
C	−1.7929 **	(0.0272)	−6.8490 **	(0.0310)
Threshold	4.9853		4.9826
SSR	102.3113		648.3256
Regime	2		2
Serial Correlation	0.8808		0.8675
Heteroscedasticity	0.5466		0.5414
Normality Test	0.1345s		0.1389

Source: author’s computations. Note: The probability values are in parenthesis while the coefficients are not in parenthesis. ‘***’, ‘**’, and ‘*’ are significance at 1 percent, 5 percent and 10 percent, respectively.

, the TAR model showed that the monetary policy rate in the prior period had significant impacts on the policy rule of the central bank, while exchange rates had significant effects on the central bank’s reaction function. This implies that it is important to accommodate previous monetary policy decisions and their performance when setting the monetary policy rate. According to a priori analysis, the implication of a positive output gap implies that the economy may be overheating, while a negative output gap means that the economy is underperforming. From our results, the direct impact of the output gap on the monetary policy rule suggests that the monetary authorities can increase interest rates to keep the economy from overheating. This outcome is plausible since a rise in the inflation gap implies that the central bank can raise interest rates to counteract inflation, discouraging excessive spending and borrowing. In addition, a threshold value of roughly 5 percent was established in both the TAR and STR models.

The STR model showed that the exchange rate is significant within the CBN’s reaction function. In the linear model, the exchange rate positively affects the CBN’s reaction function, while it also affects it positively but to a higher degree in the non-linear model (or upper regime), thereby validating the non-linear effect of a depreciation on the reaction function of the central bank in accordance with the Wald test results shown in Table 5. Consequently, the central bank of Nigeria should prioritize exchange rate depreciation policies within the threshold when formulating and implementing monetary policy decisions to ensure the effective implementation of monetary policy. In essence, the findings showed that the consideration of exchange rate management is important in the policy decisions of the central bank authorities. Finally, the results of the post-diagnostic tests using the Breusch-Godfrey serial correlation test, Jarque-Bera normality tests, and the heteroscedasticity tests using the Breusch-Pagan-Godfrey tests showed that the models are free from serial correlation, are normally distributed, and are not heteroscedastic.

4.3. Sub-Sample Analysis (2000–2015) and a Test for the Sensitivity of the Results

Based on the analysis of the main results, this study further adopted two sub-sample analyses to confirm the consistency of the main results. The first sub-sample analysis was for the period from Q1 2000 to Q4 2015. This was the period when Nigeria experienced higher levels of growth rates compared to the period from 2016 to 2023. We exempted the period of 2016 to 2023 in the first sub-sample analysis because, during these years, Nigeria experienced two separate recessions, while the economy was severely affected by the COVID-19 pandemic and Russia’s war in Ukraine in the later part of those years. The belief here is that such occurrences may affect the results more significantly. In addition, the second sub-sample period was from the period of the global financial crisis in 2008 to 2023. The reason for using these periods is to understand whether the periods of global financial crisis had any debilitating consequences on the Nigerian economy or if they would have shifted the threshold levels beyond which an exchange rate depreciation would affect prices in Nigeria.

The sub-sample analysis results on the pass-through effect of exchange rate on consumer prices, producer prices, import prices, and export prices were not significantly different from what was obtained in the main analyses using both the TAR and STR models. Therefore, these results are robust, since it shows that regardless of the period investigated, the threshold level of exchange rate depreciation on prices is roughly 5 percent. Below this level, the exchange rate pass-through would have a lower significant effect on prices, but above this level, exchange rates tend to affect prices to a greater degree, leading to greater inflation. The results for the exchange rate pass-through on consumer and producer prices are presented in

Table 9. The sub-sample threshold results – exchange rate pass-through on consumer and producer prices (Q1 2000–Q4 2015).

	Exchange Rate Pass-Through on Consumer Prices				Exchange Rate Pass-Through on Producer Prices
	Threshold Regression		Smooth Threshold Regression		Threshold Regression		Smooth Threshold Regression
	Low Regime	High Regime	Linear Model	Non-Linear	Low Regime	High Regime	Linear Model	Non-Linear
LEXC	0.7143	0.8104	0.7316	0.8392	0.7102	1.4889	0.7334	1.5189
	(0.0017) ***	(0.0000) ***	(0.0089) ***	(0.0000) ***	(0.0003) ***	(0.0019) ***	(0.0095) ***	(0.0000) ***
	Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors
LEXC(-1)	0.0351 **	(0.0171)	0.0281 ***	(0.0123)	0.1659 *	(0.0876)	0.1425 *	(0.0802)
MPR	−0.0358 ***	(0.0015)	−0.0341 ***	(0.0007)	−0.0821 **	(0.0121)	−0.0667 **	(0.0166)
MPR(-1)	−0.0245 **	(0.0388)	−0.0274 *	(0.0798)	−0.0398 **	(0.0465)	−0.0398 **	(0.0328)
RGDPG	0.0012 ***	(0.0078)	0.0059 ***	(0.0098)	0.0090 ***	(0.0061)	0.0093 ***	(0.0094)
RGDPG(-1)	0.0178 ***	(0.0054)	0.0066 **	(0.0189)	0.0059 ***	(0.0066)	0.0089 ***	(0.0083)
C	−0.8779 *	(0.0998)			2.1999 **	(0.0133)
Threshold	5.008		5.0046 ***	(0.0000)	5.0095		5.0088 ***	(0.0000)
SSR	1.1804		1.8970		11.9886		11.7863
Regime	2		2		2		2
Serial Correlation	0.5212		0.8747		0.9565		0.4799
Heteroscedasticity	0.3664		0.8951		0.9384		0.8697
Normality Test	0.8983		0.6849		0.7886		0.8686

Source: author’s computations. Note: The probability values are in parenthesis while the coefficients are not in parenthesis. ‘***’, ‘**’, and ‘*’ are significance at 1 percent, 5 percent and 10 percent, respectively.

, while the results of an exchange rate pass-through on import and export prices are presented in

Table 10. The sub-sample threshold results of exchange rate pass-through on export and import prices (Q1 2000–Q4 2015).

	Exchange Rate Pass-Through on Export Prices				Exchange Rate Pass-Through on Import Prices
	Threshold Regression		Smooth Threshold Regression		Threshold Regression		Smooth Threshold Regression
	Low Regime	High Regime	Linear Model	Non-Linear	Low Regime	High Regime	Linear Model	Non-Linear
LEXC	0.6319 ***	1.5722 ***	0.6319 ***	1.6059 ***	0.7716 ***	1.6500 **	0.7696 **	1.6490 ***
	(0.0057)	(0.0085)	(0.0001)	(0.0000)	(0.0091)	(0.0308)	(0.0426)	(0.0000)
	Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors
LEXC(-1)	0.2720 **	(0.0471)	0.8665 *	(0.0642)	0.1440 *	(0.0665)	0.2381 *	(0.0518)
MPR	−0.0075 **	(0.0481)	−0.0223 **	(0.0205)	−0.0180 **	(0.0135)	−0.0239 *	(0.0734)
MPR(-1)	−0.0137 **	(0.0324)	−0.0446 **	(0.0176)	−0.0181 **	(0.0147)	−0.0268 **	(0.0474)
RGDPG	0.0060 **	(0.0223)	0.0167 **	(0.0127)	0.0016 *	(0.0719)	0.0004 *	(0.0926)
RGDPG(-1)	0.0017 **	(0.0286)	0.0017 *	(0.0797)	0.0041 **	(0.0347)	0.0028 *	(0.0562)
C	3.1965 ***	(0.0000)	3.1523 ***	(0.0000)	1.6324 ***	(0.0058)	1.6459 ***	(0.0051)
Threshold	5.0072		5.0152 ***	(0.0000)	5.0119		5.0214 ***	(0.0000)
SSR	1.6853		2.85		1.3506		3.28
Regime	2		2		2		2
Serial Correlation	0.1657		0.277		0.138		0.1267
Heteroscedasticity	0.334		0.6018		0.614		0.5476
Normality Test	0.812		0.7657		0.3681		0.6456

Source: author’s computations. Note: The probability values are in parenthesis while the coefficients are not in parenthesis. ‘***’, ‘**’, and ‘*’ are significance at 1 percent, 5 percent and 10 percent, respectively.

. On the other hand, the results on exchange rate thresholds within the CB’s reaction function using the Taylor rule type model are presented in

Table 11. Exchange rate threshold in the Taylor rule results (Q1 2000–Q4 2015).

	Threshold Regression		Smooth Threshold Regression
	Low Regime	High Regime	Linear Model	Non-Linear
LEXC	3.1356 **	5.8341 **	1.6685 ***	5.0691 ***
	(0.0127)	(0.0189)	(0.0013)	(0.0000)
	Non-Threshold Regressors		Non-Threshold Regressors
INFGAP	0.0111 *	(0.0789)	0.2011 **	(0.0209)
YGAP	0.0259 *	(0.0521)	0.0811 **	(0.0170)
MPR(-1)	−0.9034 ***	(0.0000)	−0.2898 **	(0.0278)
LEXC(-1)	3.0016 **	(0.0287)	1.1159 **	(0.0135)
C	−1.7911 **	(0.0296)	−6.8477 **	(0.0313)
Threshold	4.9900		4.9823
SSR	102.3134		600.3065
Regime	2		2
Serial Correlation	0.8006		0.9619
Heteroscedasticity	0.6418		0.4411
Normality Test	0.1765		0.2396

Source: author’s computations. Note: The probability values are in parenthesis while the coefficients are not in parenthesis. ‘***’, ‘**’, and ‘*’ are significance at 1 percent, 5 percent and 10 percent, respectively.

. The serial correlation tests, heteroscedasticity tests, and normality tests showed that these models were normal and devoid of heteroscedasticity problems or serial correlation problems.

Finally, the sub-sample analysis results from Q1 2008 to Q4 2023 were in line with the main results and those of the first sub-sample analysis (Q1 2000–Q4 2015). As a result, these findings were robust when the sensitivity and robustness of the results were tested across a different time frame. Therefore, the findings show that exchange rate pass-through at roughly 5 percent or below would have a milder effect on prices compared to when the depreciation exceeds 5 percent, as it tends to affect prices to a greater extent. This means that over the long term, the optimum threshold level of an exchange rate depreciation in any quarter should not exceed roughly 5 percent. When it does, it has more debilitating consequences on prices in Nigeria. These findings of an exchange rate pass-through on consumer and producer prices are presented in Appendix A,

Table A1. The sub-sample threshold results of exchange rate pass-through on consumer and producer prices (Q1 2008–Q4 2023).

	Exchange Rate Pass-Through on Consumer Prices				Exchange Rate Pass-Through on Producer Prices
	Threshold Regression		Smooth Threshold Regression		Threshold Regression		Smooth Threshold Regression
	Low Regime	High Regime	Linear Model	Non-Linear	Low Regime	High Regime	Linear Model	Non-Linear
LEXC	0.7322	0.8112	0.7299	0.8298	0.7117	1.4819	0.7398	1.5217
	(0.0046) ***	(0.0000) ***	(0.0009) ***	(0.0000) ***	(0.0002) ***	(0.0009) ***	(0.0059) ***	(0.0000) ***
	Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors
LEXC(-1)	0.0360 **	(0.0178)	0.0280 ***	(0.0125)	0.1596 *	(0.0820)	0.1423 *	(0.0800)
MPR	−0.0355 ***	(0.0018)	−0.0340 ***	(0.0009)	−0.0844 **	(0.0114)	−0.0671 **	(0.0167)
MPR(-1)	−0.0249 **	(0.0390)	−0.0278 *	(0.0786)	−0.0391 **	(0.0499)	−0.0304 **	(0.0333)
RGDPG	0.0019 ***	(0.0081)	0.0063 ***	(0.0096)	0.0092 ***	(0.0056)	0.0097 ***	(0.0099)
RGDPG(-1)	0.0181 ***	(0.0045)	0.0069 **	(0.0191)	0.0061 ***	(0.0099)	0.0090 ***	(0.0088)
C	−0.8008 *	(0.0993)			2.2000 **	(0.0139)
Threshold	5.008		5.0059 ***	(0.0000)	5.0097		5.0086 ***	(0.0000)
SSR	1.1818		1.8990		11.9899		11.7899
Regime	2		2		2		2
Serial Correlation	0.5008		0.7191		0.9100		0.6709
Heteroscedasticity	0.3684		0.6950		0.9956		0.6971
Normality Test	0.8912		0.6865		0.7877		0.8610

Source: author’s computations. Note: The probability values are in parenthesis while the coefficients are not in parenthesis. ‘***’, ‘**’, and ‘*’ are significance at 1 percent, 5 percent and 10 percent, respectively.

, while the results of an exchange rate pass-through on import and export prices are presented in Appendix A,

Table A2. The sub-sample threshold results of exchange rate pass-through on export and import prices (Q1 2008–Q4 2023).

	Exchange Rate Pass-Through on Export Prices				Exchange Rate Pass-Through on Import Prices
	Threshold Regression		Smooth Threshold Regression		Threshold Regression		Smooth Threshold Regression
	Low Regime	High Regime	Linear Model	Non-Linear	Low Regime	High Regime	Linear Model	Non-Linear
LEXC	0.6319 ***	1.5722 ***	0.6319 ***	1.6059 ***	0.7716 ***	1.6500 **	0.7696 **	1.6490 ***
	(0.0057)	(0.0085)	(0.0001)	(0.0000)	(0.0091)	(0.0308)	(0.0426)	(0.0000)
	Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors		Non-Threshold Regressors
LEXC(-1)	0.2720 **	(0.0471)	0.8665 *	(0.0642)	0.1440 *	(0.0665)	0.2381 *	(0.0518)
MPR	−0.0075 **	(0.0481)	−0.0223 **	(0.0205)	−0.0180 **	(0.0135)	−0.0239 *	(0.0734)
MPR(-1)	−0.0137 **	(0.0324)	−0.0446 **	(0.0176)	−0.0181 **	(0.0147)	−0.0268 **	(0.0474)
RGDPG	0.0060 **	(0.0223)	0.0167 **	(0.0127)	0.0016 *	(0.0719)	0.0004 *	(0.0926)
RGDPG(-1)	0.0017 **	(0.0286)	0.0017 *	(0.0797)	0.0041 **	(0.0347)	0.0028 *	(0.0562)
C	3.1965 ***	(0.0000)	3.1523 ***	(0.0000)	1.6324 ***	(0.0058)	1.6459 ***	(0.0051)
Threshold	5.0072		5.0152 ***	(0.0000)	5.0119		5.0214 ***	(0.0000)
SSR	1.6853		2.85		1.3506		3.28
Regime	2		2		2		2
Serial Correlation	0.1657		0.277		0.138		0.1267
Heteroscedasticity	0.334		0.6018		0.614		0.5476
Normality Test	0.812		0.7657		0.3681		0.6456

Source: author’s computations. Note: The probability values are in parenthesis while the coefficients are not in parenthesis. ‘***’, ‘**’, and ‘*’ are significance at 1 percent, 5 percent and 10 percent, respectively.

. Lastly, the report on exchange rate pass-through within the Taylor-type monetary policy rule is presented in Appendix A,

Table A3. Exchange rate threshold in the Taylor rule results (Q1 2008–Q4 2023).

	Threshold Regression		Smooth Threshold Regression
	Low Regime	High Regime	Linear Model	Non-Linear
LEXC	3.1984 **	5.8567 **	1.70016685 ***	5.0711 ***
	(0.0131)	(0.0192)	(0.0089)	(0.0000)
	Non-Threshold Regressors		Non-Threshold Regressors
INFGAP	0.0129 *	(0.0790)	0.2000 **	(0.0289)
YGAP	0.0279 *	(0.0511)	0.0800 **	(0.0177)
MPR(-1)	−0.9089 ***	(0.0000)	−0.2678 **	(0.0212)
LEXC(-1)	3.0020 **	(0.0299)	1.1201 **	(0.0198)
C	−1.7978 **	(0.0311)	−6.8496 **	(0.0333)
Threshold	4.9918		4.9732
SSR	102.3137		600.3189
Regime	2		2
Serial Correlation	0.8770		0.9543
Heteroscedasticity	0.6657		0.4192
Normality Test	0.1908		0.2749

Source: author’s computations. Note: The probability values are in parenthesis while the coefficients are not in parenthesis. ‘***’, ‘**’, and ‘*’ are significance at 1 percent, 5 percent and 10 percent, respectively.

.

5. Discussion of Results

This study examined the effects of an exchange rate pass-through on consumer prices, producer prices, import prices, and export prices using the TAR and STR estimation techniques. The study was proposed based on the monetary perspective and the role of the monetary authorities in stabilizing exchange rates and countering inflation using monetary policy instruments. Based on the objectives of the study, six different analyses were constructed using five different models in each analysis, bringing a total of 30 regression models. These six different analyses include the main analysis using the TAR and STR methods, the sub-sample analysis from Q1 2000 to Q4 2015 using the TAR and STR techniques, and the sub-sample analysis from Q1 2008 to Q4 2023 using the TAR and STR estimation techniques. The five different models considered were the effects of exchange rate depreciation on the consumer price model, the producer price model, the import price model, the export price model, and the Taylor rule-type model.

The findings show that exchange rate depreciation has an asymmetric effect within the consumer price model, import price model, producer price model, and export price model. The asymmetric effect of exchange rate depreciation on the import price model is in line with some previous studies, such as Adekunle et al. (2019) in Nigeria and Auer and Schoenle (2016) in the US, while its asymmetric effect on producer prices corroborates a previous study by Choudhri et al. (2005). On the other hand, its asymmetric effect on consumer prices is similar to some previous studies on Nigeria, such as Adekunle et al. (2019) and Bada et al. (2016), and some other studies abroad, such as Aleem and Lahiani (2014) for the East Asian Countries and Latin America and Oktay (2022) for Switzerland, while the effects of currency depreciation on export prices were in line with Oyinlola and Babatunde (2009) for Nigeria.

The findings of the TAR model showed that the threshold level beyond which currency depreciation may have more debilitating effects on consumer prices, producer prices, import prices, and export prices is roughly 5 percent across these four models. Furthermore, by employing the Taylor rule on the nexus between currency fluctuations, inflation, and monetary policy decisions, the study also found a threshold of 5 percent, beyond which a currency depreciation will harm monetary policy decisions and the economy at large. The effect of currency depreciation within the Taylor rule is in line with previous findings, such as Caporale et al. (2018), Ozdemir (2020), and Valogo et al. (2023). The sub-sample analysis results from Q1 2000 to Q4 2015 and from Q1 2008 to Q4 2023 also corroborate the main findings of a 5 percent threshold in any quarter, beyond which an exchange rate depreciation may hurt prices in Nigeria. The results are consistent across the four different models (consumer price model, producer price model, import price model, and export price model), revealing that the results are robust regardless of the variable used to measure prices or inflation in Nigeria.

The findings of the STR model are in line with the findings of the TAR model. This means that, overall, exchange rate pass-through has an optimum threshold of 5 percent, beyond which it hurts prices and then inflation in Nigeria. Below this threshold, an exchange rate depreciation will have a lower effect or lower pass-through on prices and inflation in Nigeria, albeit to a high degree. Above this optimum threshold, however, the effect of currency depreciation has a more pronounced debilitating consequence on prices in Nigeria, overshooting the pass-through effect on the producer price model, import price model, and export price model, while it has an extremely high pass-through on the consumer price model. The reason why the consumer price model does not overshoot may be because not all increases in the input prices or the cost of producing an item can be passed down to the consumers. However, the fact that exchange rate pass-through has a very high degree of pass-through shows that most costs borne by exchange rate depreciation are passed on to households, who are the final consumers of these items. Finally, the Taylor rule results of the STR model also found an optimum threshold of 5 percent depreciation. This finding implies that in any given quarter, for the country to remain competitive both locally and globally, exchange rate depreciation must not go beyond 5 percent, otherwise this will have more harmful consequences on the economy. The findings of this study are similar to previous studies, such as Caporale et al. (2018), Ozdemir (2020), and Valogo et al. (2023). A summary of the threshold regression results using the STR and TAR models across the different price models can be found in Figure 2.

6. Conclusions

This paper explored the threshold effect of currency adjustment on prices in Nigeria by determining the effects of a currency depreciation pass-through on producer prices, consumer prices, export prices, import prices, and the monetary policy reaction function using the Taylor rule model during the period from 2000 to 2023, using quarterly data. To obtain the results, the study adopted threshold regression and the self-exciting smooth threshold regression procedures to confirm the consistency in the threshold results and account for the lower and upper threshold estimations. Overall, the findings established that exchange rate pass-through affects prices in Nigeria in different ways below and above the threshold value using both threshold regression methods.

The findings established a threshold of 5 percent for a quarterly currency depreciation on prices (consumer prices, producer prices, export prices, and import prices) using the TAR and STR models. This demonstrates that an average yearly depreciation not higher than 20 percent is reasonable if Nigeria is to remain competitive in the global international trade market, and if domestic firms want to remain competitive while maintaining price pressures. Therefore, a depreciation will weaken the domestic currency against the international benchmark, leading to a rise in import prices and thereby raising consumer and producer prices. Finally, the results suggest that the currency adjustment pass-through to producer prices, consumer prices, export prices, and import prices are considerably higher in Nigeria below the threshold. Above the threshold, however, it overshoots for producer prices, export prices, and import prices, which are the reasons for spiraling inflation in the economy. The sub-sample analysis from Q1 2000 to Q4 2015 and from Q1 2008 to Q4 2023 were in line with the main analysis. This implies that regardless of the periods investigated, the optimum exchange rate threshold on prices does not change in Nigeria.

Furthermore, a threshold value of roughly 5 percent exchange rate depreciation was established within the central bank’s reaction function, both in the main analysis and the sub-sample results. In addition, the inflation and output gap significantly affected the monetary policy rule using Taylor’s equation. This outcome is plausible, since the monetary authorities were not within these threshold limits of average inflation and output targets over the period of study as established in the findings. This result implies that the Central Bank of Nigeria should prioritize exchange rate depreciation policies when designing and implementing monetary policy decisions to ensure the effective implementation of monetary policy. In addition, the currency depreciation within the Taylor rule is lower below the threshold than above the threshold. In essence, the findings showed that the inclusion of exchange rate policies is extremely vital when designing monetary policies.

To keep inflation in check, this paper suggests that the monetary authorities should try to keep exchange rate depreciation below the established thresholds. This can be achieved by intervening in the foreign exchange market through the adoption of foreign exchange forwards by infusing sufficient foreign currency to meet rising demand. This will, however, depend on the level of foreign reserves available to cushion the effect of the foreign currency demand. In addition, the government must create an enabling environment to help businesses, such that existing producers can rely more on domestic organizations for their factor inputs to curtail the effect of external events on domestic prices. Backward integration and import substitution policies might also be adopted. This would help reduce the overdependence on imports of factor inputs and strengthen the domestic currency against the international benchmark. Furthermore, the monetary authorities should consider adjusting the policy rate to take into account exchange rate depreciation thresholds to keep domestic prices stable.

In addition, the findings have reinvigorated the argument for the monetary authorities of Nigeria’s Central Bank to fully adopt an inflation-targeting strategy to help combat inflation and stabilize prices in Nigeria. Moreover, implementing a credible inflation targeting strategy may help anchor inflation expectations, making the negative effects of exchange rate fluctuations less likely to influence price stability. To achieve this, the Central Bank must ensure that its monetary policy is transparent, predictable, and focused on keeping inflation in check, while also diversifying its sources of foreign exchange; otherwise, its over-reliance on oil and a few other sources may continue to have an effect when there are shocks affecting oil demand and supply or prices in the global crude oil market.

Last but not least, another suitable solution for the current exchange rate depreciation problem in addition to the inflation targeting strategy could be to adopt a ‘crawling peg’ exchange rate framework to effectively manage the exchange rate crisis in Nigeria. A crawling peg is an exchange rate adjustment method in which a currency with a managed float exchange rate is permitted to vary within a band rate. As a result, adopting a crawling peg for Nigeria would mean setting the country’s official exchange rate at or near the true market value of the Naira per U.S. dollar (NGN/USD), and allowing the Naira to depreciate or appreciate against the USD by roughly the threshold of 5 percent. This will help businesses and households plan their consumption activities better if the monetary authorities are more transparent in their dealings. The limitations of this study stem from the fact that quarterly data could only be gathered from the year 2000 to 2023 to understand this dynamic. However, future studies may look at how to suppress the negative impact of exchange rate depreciation over the West African corridor to extend this discussion on how exchange rates affect prices in the West African region, since the region is predominantly import-dependent.