Modelling intervention of Columbian peso to Nigerian naira exchange rates due to 2016 & 2020 Nigerian economic recessions

Introduction

Nigeria is primarily dependent on revenue generated from oil exports, rendering it susceptible to the fluctuations and instability inherent in the global oil market. The year 2016 was characterised by a notable economic downturn in Nigeria, primarily due to a substantial decrease in oil prices. This reduction in oil prices resulted in considerable budgetary difficulties and a contraction of the economy. The regulation of exchange rates emerged as a crucial element in these actions. Foreign exchange rate policies play a crucial role in countries such as Nigeria, as they exert influence over the competitiveness of the nation's exports and imports, exert an impact on inflation rates, and has the potential to affect the general economic stability. In periods of economic turmoil, central banks frequently employ interventions in the foreign exchange market as a means to regulate the valuation of their currency. The decision to prioritise the Colombian Pesos as currency of importance in this analysis is presumably motivated by their significance as trading partners with Nigeria. Colombia possesses substantial economic connections with Nigeria, characterised by the exchange and influx of trade and investment. An examination of the impact of interventions in the exchange rates of various currencies on the Nigerian Naira can provide insights into the efficacy of such policies in times of economic downturns. Similarly, the analysis is grounded in the backdrop of Nigeria's susceptibility to external shocks in its economy, specifically the variations in oil prices, and the difficulties presented by the collapse in oil prices in 2016. Thus, this work focused on examining the fluctuation in the relative worth of the Nigerian Naira in relation to the Columbian Peso (COP) during the economic downturn of 2016. In past study, Mosugu and Anieting (2016) [9] employed intervention analysis as a methodological framework to evaluate the effects of governmental regime and policy alterations on foreign currency rates within the Nigerian context. Shittu (2009) [13] utilised intervention analysis as a methodological approach to examine the monthly variations in exchange rates between the Naira and the US Dollar within the time frame of 1970 to 2004.

The researcher successfully identified various intervention components during the course of their investigation. In recent work on intervention, Inyang et al (2023) [6] worked on Time Series Intervention Modelling Based on ESM and ARIMA Models: Daily Pakistan Rupee/Nigerian Naira Exchange Rate. Etuk et al (2022) [3] , investigated the impact of declaration of cooperation (DoC) on the Nigerian crude oil production. Moffat and Inyang (2022) [8] , investigated the impact of the Nigerian government amnesty programme (GAP) on her crude oil production. Etuk et al (2021) [4] used Arimaintervention Analysis in modelling Nigerian Automotive Gas Oil Distribution. Etuk and Amadi (2021) [5] modelled Nigerian Monthly Crude Oil Prices using Arima-intervention model. Shittu and Inyang (2019) [12] modelled Nigerian monthly crude oil prices using the ARIMA-Intervention model with a view to comparing the result with that of the intervention model using lag operator. Wiri and Tuaneh (2019) [14] modelled the Nigerian Crude Oil Prices Using ARIMA, Pre-intervention and Post-intervention Model. Mrinmoy et al (2014) [10] used time series Intervention Modelling for Modelling and Forecasting Cotton Yield in India. Jarrett and Kyper (2011) [7] , used ARIMA Modelling with Intervention to Forecast and Analysed Chinese Stock Prices. Roy et al (2009) [11] used ARIMA -Intervention Analysis in Modelling the Financial Crisis in China's Manufacturing Industry.

Model Specification

The transfer function-noise model proposed by Box and Tiao (1975) [2] is given as

Where, 𝑍 𝑡 is the response variable at t, 𝑏 =delay parameter, 𝜔 𝑠 =impact parameter, 𝛿 𝑟 =slope parameter, ∅ =Non-seasonal autoregressive parameter, 𝜃 =Non-seasonal moving average parameter, 𝑎 𝑡 =White noise, 𝑋 𝑡 = Input function or Indicator variable

Mathematically, there exist two input functions:

Data Description

The dataset comprises daily exchange rates between the Nigerian Naira and the Columbian Peso, as well as the Mexican Peso, for the periods of January 1st to August 31st in 2016, and September 1st to December 31st in 2020. The exchange rates were obtained from the websites http://www.exchangerates.org.uk/COP-NGN-spot-exchange-rates-history-2016.html and http://www.exchangerates.org.uk/COP-NGN-spot-exchange-rates-history-2020.html. The research was conducted with EViews statistical software packages.

Results and Discussions Analysis of the 2016 Columbian Peso to Nigerian Naira Exchange Rate

The data used for the analysis covers the period from January 1st to August 31st, 2016. The result in Table 1 favored the null hypothesis that the preintervention series has a unit root (i.e. that the preintervention series is non-stationary) since the probability value of 0.4731 is greater than 0.05.

Since the pre-intervention series is non-stationary the series was differenced and the time plot of the order 1 differenced pre-intervention series is given in Figure 3. From Table 2 the probability value of 0.000 implies that the pre-intervention series becomes stationary after differencing.

Again, the correlogram of the differenced pre-intervention series given in Table 3 indicates a white noise fit for the series. 3 shows that the fitted model is a white noise model since F 1 = 0.0686 Having established the stationarity of pre-intervention series and a white noise fit for series the transfer function for the intervention model for the 2016 COP/NGN exchange rate series was obtained as shown in Table 4. The intervention model Z given in Table 4 where C(1) and C(2) are the coefficients and T is time after the series started, was used to make an intervention forecast of the 2016 post-intervention COP/NGN exchange rate and the graph of the superimposition of the forecast to the observed data is given in Figure 4.

Analysis of the 2020 Columbian Peso to Nigerian Naira Exchange Rate

The data used for the analysis covers the period from September 1st to December 31st, 2020. Figure 5 depicts a time plot illustrating the daily exchange rates between the Columbian Peso (COP) and the Nigerian Naira (NGN) from September 1st to December 31st, 2020. On the other hand, Figure 6 presents a time plot showcasing the daily COP/NGN exchange rates specifically within the preintervention period of September 1st to November 19th, 2020, consisting of 80 recorded rates. From Figure 5 it can be deduced that the daily COP/NGN exchange rates collected are not stationary. The probability value of 0.5301 in Table 6 leads to the acceptance of the null hypothesis that the 2020 daily COP/NGN pre-intervention exchange rates collected has a unit root.

Therefore, the daily COP/NGN pre-intervention exchanged rates were differenced to achieve stability and the time plot given in Figure 7. The unit root test of the differenced 2020 daily COP/NGN pre-intervention exchange rate given in Table 7 shows that the series became stationary after differencing based on the probability value of 0.0001 obtained. The correlogram in Table 8 indicates that the differenced 2020 daily COP/NGN pre-intervention exchange rate produced a white noise fit. Given that the differenced 2020 daily COP/NGN pre-intervention exchange rate is stationary and have a white noise fit the intervention model was determined as shown in Table 9. The intervention model Z1 given in Table 9 where C(3) and C(4) are the coefficients and T is time after the series started was used to forecast the post intervention 2020 daily COP/NGN and the forecast was superimposed on the 2020 daily COP/NGN post-intervention exchange rate is given in Figure 8.

Results and Discussion

Non-stationarity is a prevalent characteristic observed in the majority of economic variables. The exchange rates exhibited non-stationarity, as evidenced by the data presented in Figures 1 and5

Conclusion

This study examines the fluctuations of the Columbian Peso (COP) in relation to the Nigerian Naira (NGN) throughout the years 2016 and 2020. The non-stationarity of the COP/NGN exchange rates, as well as their pre-intervention series, was evident based on the observed data. However, the exchange rates exhibited stationarity after being differenced for the first time. The Chi-square goodness of fit tests was conducted to assess the significance of the mean change in exchange rates between the pre-intervention and post-intervention series, based on the forecast generated by the intervention model.

References

1. Box GEP, Tiao GC. Intervention Analysis with Application to Economic and Environmental Problems. Journal of American Statistical Association. 1975;70(349):70-79.
2. Etuk EH, Inyang EJ, Udoudo UP. Impact of Declaration of Cooperation on the Nigerian Crude Oil Production. International Journal of Statistics and Applied Mathematics. 2022;7(2):165-169.
3. Etuk EH, Moffat IU, Agbam AS. Intervention Analysis of Nigerian Automotive Gas Oil Distribution. Nigerian Journal of Oil and Gas Technology, 2021, 3(2).
4. Etuk EH, Amadi EH. Arima Intervention Analysis of Nigerian Monthly Crude Oil Prices. Nigerian Journal of Oil and Gas Technology, 2021, 3(1).
5. Inyang EJ, Etuk EH, Nafo NM, Da-Wariboko YA. Time Series Intervention Modelling Based on ESM and ARIMA Models: Daily Pakistan Rupee/Nigerian Naira Exchange Rate. Asian Journal of Probability and Statistics. 2023;25(3):1-17. Article no.AJPAS.106693. ISSN: 2582-0230. DOI: 10.9734/AJPAS/2023/v25i3560
6. Jarrett JE, Kyper E. ARIMA Modelling with Intervention to Forecast and Analyze Chinese Stock Prices. Int. J. Eng. Bus. manag. 2011;3:53-58.
7. Moffat IU, Inyang EJ. Impact Assessment of Gap on Nigerian Crude Oil Production: A Box-Tiao Intervention Approach. Asian Journal of Probability and Statistics. 2022;17(2):52-60. https://doi.org/10.9734/ajpas/2022/v17i230419
8. Mosugu JK, Anieting AE. Intervention Analysis of Nigeria's Foreign Exchange Rate, Journal of Applied Science and Environmental Management, 2016;20(3):891-894
9. Mrinmoy R, Ramasubramanian V, Amrender K, Anil R. Application of Time Series Intervention Modelling for Modelling and Forecasting Cotton Yield. Statistics and Applications. 2014;12(1&2):61-70.
10. Roy CPC, Ip WH, Chan SL. An ARIMA -Intervention Analysis Model for the Financial Crisis in China's Manufacturing Industry. Intervention Journal of Engineering Business Management, 2009, 1(1).
11. Shittu OI, Inyang EJ. Statistical Assessment of Government's Interventions on Nigerian Crude Oil Prices. A Publication of Professional Statisticians Society of Nigeria, Proceedings of 3rd International Conference. 2019;3:519-524.
12. Wiri L, Tuaneh GL. Modelling the Nigeria Crude Oil Prices Using ARIMA, Pre-intervention and Postintervention Model. Asian Journal of Probability and Statistics. 2019;3(1):1-12. Article No. AJPAS.46900n