Commodity Price Shocks And Macroeconomic Performance In Sub-Saharan Africa

This study investigated the relationship between commodity price shocks and output growth in Sub-Saharan African countries using panel data which covered the period between 2005 and 2017. Data for the study were sourced from the World Development Indicators (WDI). Data base of the World Bank, The IMF’S International Financial Statistics (IFM) and Publications of Central Banks of various countries selected. The study employed Generalized Movement Average (GMM) as the estimation technique. Findings from the study showed that positive changes in the prices of export commodities has little positive impact on macroeconomic performance in Sub-Saharan Africa while negative price change has negative and significant impact on macroeconomic performance in Sub-Saharan African countries during study period. Based on these findings, the study therefore concludes that the relationship between commodity price shocks and macroeconomic performance in Sub-Saharan Africa is asymmetric. The study recommends that countries in Sub-Saharan Africa should introduce and implement policies to withstand shocks that may come from commodity price shocks such as economic diversification not only in area of agriculture but also in the area of industrialization and manufacturing.


SECTION ONE Introduction
Fluctuations in the commodity prices cut across both oil and non-oil producing countries in Africa. For instance, the sudden reduction in crude oil price at international oil market exposed oil producing countries particular Africa oil producing countries to various degrees of challenges in managing their macroeconomic performance particularly Nigeria, Ghana, Cameron, Angola, Equatorial Guinea, South Sudan, Sudan, Republic of the Congo, Gabon and Chad while non-oil producing states such as Samba, Zimbabwe, Senegal, Tanzania experienced slow economic growth because of agricultural commodity price fluctuations. Other African countries have their own fair share from severe drought.
Fluctuation and reduction in production and exploitation of certain commodities (Oil and Agricultural products) came in recent time as a result of various attacks by some groups. Take for instance, Islamic extremists, such as Boko-Haram in Nigeria, Niger, Cameroon, Siea in Sudan, Islamic fundamentalist in Central Africa and also the Niger delta militant in Nigeria. These fluctuations in commodity price might have influenced the macroeconomic performance of Sub-Saharan African economies either positively or negatively. It has been established that commodity price uncertainty and shocks pose more challenges for those nations that engage in the production and exportation of primary and intermediate products, World Bank (2014). These challenges are experienced not only by government, also producers in the affected countries have their own fair share. Take for instance, envisaged changes in export prices may complicate budgeting process and can hamper the attainment of debt targets. This is very common in developing nations that mostly depend on earnings from export of primary and intermediate products for buck of their revenues. For producers (exporters) price fluctuation stimulates increase in cash flow variability which eventually retards collateral value of inventories, this combine to accelerate interest rate for peasant producers. Also, inadequate savings and lack of access to credit facilities cope with revenue fluctuation through crop diversification may likely ignore specialization which may eventually leads to a roning stone that garders no moose. (hack of specialization).
Furthermore, commodity price uncertainty may generate inflationary pressure especially in developing nations (Sub-Saharan African inclusive). Majority of the countries in Sub-Saharan Africa might have had their economic growth been influenced positively or negatively, this might have occurred as a result of unstable commodity prices. Inflation as it has been established by some studies Saliu (2004) Muham, (2000) and Godwin, (2007) is not only being driven by oil price changes but also determined to some extent by food prices. A good example is the global food price increased in 2008 especially for wheat, soya-bean and corn. Domestic commodity prices can as well be driven by global commodity prices. The identified determinants of increase in commodity prices are gab between supply and demand for foods, the growing presence of financial investors and speculator in the commodity markets in case of financialization of commodities. Also, unpredictability of weather is another major reason for decrease or increase in commodity production and prices.
Additionally, in African countries, particularly in Sub-Saharan Africa, the relationship between commodity and macroeconomic performance is worrisome and paradoxical. Paradox in that, when there is an increase in commodity prices, particularly oil price, the macroeconomic performance mostly remains constant or relative increase, but when prices of these commodities fall at international markets, macroeconomic performance suffers greatly. A very good example was the recent reduction in crude oil price of mid-2014. During this period, many of the African oil producing countries were exposed to different macroeconomic instability to an extent that some of their economies went into recession.
There have been studies on the relationship between commodity price fluctuation and macroeconomic variables, Michael (2014), Jan (2000), Stephen, (2014), Paul and Benedict, (2007), Afolabi (2017) and others. However, empirical findings from most of these studies produced conflict results which might be due to the fact that some fundamental variables (investment and domestic interest rate) were not incorporated into their models. Most of the previous studies concentrated on prices of commodity rather than nature and pattern of unstable commodity price which is one of the major determinants of macroeconomic performance. This therefore, makes this present studies to depart relatively from some of the previous studies in two fold. Firstly, in this study, macroeconomic performance was measured by two indicators; namely, investment and output growth rate and also the trends of major agricultural products which form the bulk of a export goods in Sub-Saharan Africa countries were considered.
The rest of the paper is structured thus; this introductory section is followed by section two that presents the empirical literature. Section three centers on methods and materials. Section four deals with results and discussions while section five concludes the paper.

SECTION TWO Empirical Literature
Studies have been conducted both in developed and developing countries to consider the relationship between commodity prices and macroeconomic variables. Some of these studies are presented here empirically to provide guides and directions for this present studies.
Tony and Alanu (2014) examined the agriculture commodity price shocks and their effects on growth in Sub-Saharan Africa. The study employed vector autoregressive model as estimation technique. Findings from the study revealed that there was evidence of asymmetric responses of growth of selected economies to shocks emanating from commodity price shocks. In a similar study, Michael (2014) investigated the impact of commodity price shock (copper and chile) on the growth of selected economies in Africa using VAR. Findings from the study showed that both the negative and positive commodity price fluctuation impacted the economies of the selected countries. The negative commodity price has much impact on the selected economies. Cashin, (2002), employed VAR to estimate the relationship between commodity price fluctuation and output growth. A comparative study of oil and non oil producing countries. Finding showed that commodity price fluctuation has much negative impact in oil producing countries than non oil exporting countries. Frankel and Campell, (2006), studied the connections between monetary policy in agricultural sector and mineral commodity in Liberia. The study employed GARCH as estimation technique. Finding from this study revealed that monetary policy showed a considerably impact on the selected two sectors of Liberian economy. Spatafora and Samke, (2012) examined the impact of commodity prices shocks on fiscal outcomes in some selected African countries. The study employed vector autoregressive as estimation technique. Finding from this study showed that government revenue and expenditure response to commodity price increase. Arezki and Bruckerner, (2010) investigated the effects of international commodity price shocks on external debt in ninety three countries of developing nations. The study employed the conventional methods of structural Vector Autoregressive model as estimation technique. Finding revealed that increase in the international commodity price for export commodity were associated with significant reduction in external debt. In another study, Stephen, (2014), examined the sectoral impact of commodity price shocks in Australia. The study employed structural vector autoregressive as estimation techniques. Findings from the study revealed that commodity price shocks largely affect the mining, construction and manufacturing industries in Australia. However, the financial and insurance sectors were discovered to be relatively affected. Jan (2000) studied commodity price uncertainty and shocks in developing countries. The study employed GMM as estimation technique. Findings showed that per-capital growth rates are significantly reduced by large discrete negative commodity price shocks. Walter and Alfred, (1990) investigated commodity price fluctuations and macroeconomic adjustment in some selected developing countries. The study made used of Granger causality as estimation technique. Findings showed that commodity price fluctuation brought about instability in macroeconomic performance of those selected developing countries. In a related study, Nicholas, (2016) examined commodity price and income inequality from 2000 to 2014 in selected developing countries using panel co-integration. Finding from these studies showed that there was no significant relationship between commodity price and income inequality in the selected countries.
Jing He (2009) examined macroeconomic response to crude oil price change in China between 1999 and 2004. The study employed input-output as estimation technique. Finding from the study revealed that the response of macroeconomic fundamentals to shocks emanating from oil price change was positive and significant. In the same line of study, Landgraf and Chowdhury (2010) investigated the relationship between global liquidity and commodity price in emerging economies between 1980 and 1998. The study made use of Structural vector autoregressive distributive model as estimation technique. Finding from the study showed that excess liquidity was found to be one of the major determinants of growth and variation in prices than unexpected demand shocks during the study period. Thomas (2016) examined the impact of commodity price uncertainty on the growth of developing countries using panel cointegration. Finding from this study showed that commodity price change has no significant impact on the growth of developing countries. In an attempt to advance literature on this, Hassan and Salim, (2011), examined the interactions among commodity prices, inflation, unemployment and short term interest rate in Australia between 1980 and 2009. The study employed VAR as estimation technique. The results from the study revealed that positive shocks from commodity price stimulate both inflation and unemployment during the study period.
To further contribute to the existing literature, Hangnite, (2015) examined the effect of commodity prices on economic growth in African countries between 1999 and 2014. The study employed panel Structural VAR as estimation technique. Result from this study showed that an increase in commodity price is more likely to benefit African economies than hurting them. Also, Deaton and Thomas (2106), examined the relationship among commodity price shocks, growth and structural transformation in low-income countries. The study made used of panel VAR as estimation technique. Finding revealed that macroeconomic conditions do not respond to individual low income country conditions contemporaneously. This implies that a one standard deviation increase in commodity prices rises per capital income in developing countries by 26% and government spending investment by 4.4% and 12.4% conclusively.
In summary, the empirical literature reviewed above is far reaching with respect to the relationship between commodity price fluctuation and macroeconomic performance in Sub-Saharan Africa countries. However, findings from those studies reviewed produced contradictory results, Salim asymptotically. This present study tries to overcome the shortcomings in the previous studies by using quarterly data and GMM as estimation techniques.

SECTION THREE
In this section, theoretical underpinning, model specification estimation techniques and sources of data are presented.

Model Specification
The model for this study takes it root from sport price model, sticky and flexible price model that capture the seasonal effects of price dynamics and effects of money on growth and savings. However, space constraint may not permit us to illustrate this theory mathematically.
Two models are being specified to capture the relationship between commodity price shocks and macroeconomics performance in Africa. Gross fixed capital formation and real gross domestic products growth rate are used to proxy macroeconomic performance. These two variables are used as dependent variables, while price index of the export commodity, terms of trade index, real effective exchange rate and inflation rate are explanatory variables.

SECTION FOUR 4.1 Empirical Analysis
This section presents the empirical analysis of the model. The first step was to determine the behaviour of the variables by conducting the descriptive statistics. Next was the panel unit root test for the variables of interest. Here, the IM Pesaran and Shin panel test were explored. Thereafter, the trends analysis of the major export commodities were examined. Also, the interactions among the variables of interest were considered by GMM. Finally, the discussion of finding was presented. The coefficients of variation were 0.4685396, 0.0779236, 0.0448127, 0.3173405, 0.0718928, 1.1456212 and 0.093211 also respectively. Values for the GDP, export price, inflation rate and gross fixed capital formation were negatively skewed. The peakedness of the GDP and the inflation rate were higher than the normal distribution (i.e. leptokurtic) while that of the terms of trade and the gross fixed capital formation were lower (i.e. platykurtic).

Correlation Test
The pair wise correlation coefficients of all the variables employed in the study are presented in table 4.2. The result showed that there were weak correlation between gross domestic product (GDP), export commodity price (XCP). Real effective exchange rate (REER), World Oil Price (WOP), inflation rate and they were not also statistically significant, while the result showed positive and significant relationship between output growth rate, gross fixed capital formation and term oftrade. There was positive correlation between inflation (INF), GDP, export commodity prices and real effective exchange rate. Terms of trade (tot) shows that oil price has statistically significant and positive correlation with GDP, export commodity prices and inflation. Conversely, a negative and statistically insignificant correlation was obtained between terms of trade and real effective exchange rate. The relationship among gross fixed capital formations (GFCF), GDP, and export commodity prices was positive and not statistically significant. However, gross fixed capital formation exhibited a negative correlation with real effective exchange rate. The relationship between gross fixed capital formation and inflation as well as terms of trade was positive and statistically significant. Also, the relationship between oil prices was negative and also insignificant.    From the trends analysis, overview of agricultural out in countries selected across Africa region reflect a clear cut indication that Nigeria out-performed other Africa countries in agricultural production in the last 3 decades reason which could be attributed among other things to the rich agricultural land in the country as well as labour availability due to the population size of the country (7 Th Largest in the world).

System GMM Results
The result of the one-step system GMM is presented in table 4.4 where GDP is the dependent variable. Sargan test was conducted to ascertain the validity of the instrument employed in the study.
The statistically insignificant value of the Sargan statistics showed that the choice of the instrument was appropriate. Also, the model was tested for auto correlation in the first second order. Although, there was an indication of serial correlation of the error in the first model, no evidence of autocorrelation was observed in the second model. In the result obtained on the effect of export commodity prices on real GDP growth, it was reviewed that the term of trade and inflation were statistically significant at 5% and 10% levels. Other variables, namely export commodity prices, real, effective exchange rate and oil price were not statistically significant. A close look at the coefficient of parameter suggests that export commodity prices showed a positive relationship with GDP growth. It showed that a percentage increase in price of export commodity lead to GDP growth. Since terms of trade showed a positive and statistically significant relationship with GDP growth. It implies that a percentage increase in terms of trade brings 0.4% improvement in GDP growth in SSA. In line with the theory, inflation exhibited a negative relationship with GDP growth this showed that a percentage increase in inflation lead to 0.2% decline in the GDP growth in SSA. A plausible reason for the statistically insignificant value of the coefficient of export commodity prices could be as a result of Dutch Disease phenomenon that characterized export commodity in SSA. Also, world oil price showed positive but insignificant relationship with output growth. The reason for this might be as a result that world oil price is unstable in recent time especially from 2008 till date. The dynamic panel data estimation for model 2 with gross fixed capital formation as the dependent variable is present in table 4.5. Again, the validity of the instruments was tested using the Sargan test of overriding restrictions. The statistically insignificant value of the Sargan statistic showed that the instruments employed were valid. The model was tested for autocorrelation in the first and second order respectively. It was revealed that although there was serial correlation of errors in the first model, there was no evidence of autocorrelation in the second model. The estimation showed that the one period lagged gross fixed capital formation and inflation were statistically significant. This suggests that a percentage increase in previous year domestic investment would increase the current year investment by 0.9%. Contrary to theory, inflation exerted a positive and statistically significant effect on gross fixed capital formation. This implies that a one percent rise in inflation would increase gross fixed capital formation by 0.1%. The coefficient of other variables in the regression namely, export commodity prices, real effective exchange rate, oil price and terms of trade were not statistically significant. The statistically insignificant value of the coefficient of export commodity prices could be as a result of the fact that most of the proceeds from export commodity particularly oil are not being utilized appropriately. Also, statistical positive and significant relationship between oil price and investment in the model shows that to some extent oil price determines investment level in the selected Sub-Saharan African countries.

DISCUSSION OF RESULTS
The presentation and analysis result of this study commenced with the descriptive statistics which showed that all the variables used in the estimation have means that were much closed to the maximum limit. This shows that the commodity prices were relatively high but did not contribute significantly to the growth of the selected economies during the study period. The results from descriptive statistics further showed that, all the variables of interest employed in this were negatively skewed. However, the peakedness of the GDP and inflation rate were higher than the normal distribution (i.e. Leptokurtic) while that of the terms of trade and the gross fixed capital formation were lower (i.e. Platykurtic). This finding is in line with the finding of Olomola and Adejumo, (2006) as well as Olajide, (2010) where it was established that there was trade-off between imported commodity prices in developing countries particularly Nigeria. This however, always brings about unfavorable terms of trade. Also, oil price showed the highest variance among the variables of interest. The implication of this is that, oil price is highly unstable. This is clearly evident of the incessant changes in the crude oil market since 1980s. this finding is equally compatible the result obtained by Aliyu, (2011), where it was found that in the last two decades, there have been unprecedented in the growth of developing economies as a result of dynamics in oil price at international oil market. After the descriptive statistics, what followed was IM Pesaran and Shin panel unit root test. From the results, all the variable of interest became stationary after the first difference. This finding is in line with the study of Gabriel, (2016). Thereafter, trends of the most exported commodities in Sub-Saharan Africa were done. From these trends analysis, it was revealed that both upward and downward trends were followed by most of the commodities employed. This finding also corroborates the finding of Afolabi, (2016). The correction test was equally carried out in this study, it was established that there was weak correlation between gross domestic product and export commodity prices. This finding is also in line with finding of Ojo, (2011), but negates the finding of Emmanuel, (2013). Results from GMM results showed that term of trade and inflation were significant determinants of investment while other variables did not show significant relationship with investment proxed by gross fixed capital formation. However, the price of export commodity did not show significant relationship with output growth during the study period. The statistically insignificant value of the co-efficient of export commodity prices could be as a result of the fact that most of the proceeds from export commodity during booms were not effectively utilized. The finding is in line with the study of Ogunmayowa, (2016), Folorunso, (2010), but negates the finding of Gabriel, Lao, (2016).

SUMMARY AND CONCLUSION
The study examined the relationship between commodity price shocks and macroeconomic performance in Sub-Saharan Africa between 2005 and 2017. The study empirically analyzed the level of relationship between prices of exported commodities and output growth rate. In this regards, output growth rate was tested with other variables such as term of trade, oil price, real effective exchange rate, inflation Rate and the level of investments which was captured by fixed capital formation.
The results from the study showed that changes in the prices of export commodities had little though positive impact on the output growth rate. However, terms of trade had much greater influence on output growth rate while a rise in the level of inflation leads to a fall in output growth rate in those countries selected for this study. Also, the study showed that level of inflation contrary to theory displays that an increase in the level of investment in particular year could lead to further increase in the level of investment in the subsequent year. Whereas an increase in the prices of export does not lead to the much expected increase in the level of investment as confirmed in previous studies, Michael (2014), Tony and Alamu, (2014), Jan, (2000), Afolabi (2017) and others. Generally, the growth rate of output and an increased in the level of investment were associated with increased in the price of export commodities, world oil price of inflation rate as well as a favorable Terms of Trade.

POLICY IMPLICATIONS
These inferences can be drawn from the results of this study; i. That investment level in the selected countries is determined to some extent by price of export commodity, Agricultural products, Terms of Trade, Inflation rate and World Oil Price. ii. Also, an increase in the price of export commodity tends to produce favorable balance of trade thereby contributing positively to the output-growth rate in the selected countries iii. Unfavorable balance of trade reduces the level of investments. iv. Inflation neglects theory by investment captured by grossed fixed capital formation showing a positive relationship with investment.
Based on these findings, the study therefore concludes that positive changes in export commodity prices did not bring significant positive change in macroeconomic performance while negative change in price of export commodity brings negative and insignificant charges in the macroeconomic performance in Sub-Saharan Africa countries. In this regards, the study recommends that governments in various selected countries should identify the sources of commodity price shocks when considering policy option in response to commodity price