The Impact of Rand/Pula Exchange Rate Volatility on Botswana’s Economic Growth

This paper undertakes an investigation of the impact of the Rand/Pula exchange rate volatility on Botswana’s economic growth. The paper is using annual time series data, from 1977 to 2018. The Generalized Method of Moment (GMM) is employed to evaluate the impact of the real exchange rate volatility on Botswana’s economic growth. The GARCH model results found the Pula/Rand exchange rate to be volatile. The Rand/Pula exchange rate volatility does not have an impact on Botswana’s economic growth. This finding mirrors those of Kaur et al. (2019) and Musyoki et al. (2012). They found negative but insignificant impact of exchange rate volatility on economic growth in Malaysia and Kenya, respectively. Our empirical findings suggest that Botswana’s economic growth is largely explained by trade openness and growth of labour force and not influenced by the Rand/Pula exchange rate volatility.


Introduction
This paper investigates the impact of Rand/Pula exchange rate volatility on Botswana's economic growth. The Pula and the Rand are Botswana's and South Africa's currencies respectively. Exchange rates are described as the cost of foreign currency per unit of domestic currency or domestic currency per unit of foreign currency. Exchange rates allow us to express the cost or price of a good or service in a common currency (Edwards, 1988;Krugman & Obstfeld, 2000). Exchange rates are used as a form of decision rule between the domestic country and the international economy. Hence, exchange rates are key influencers of the direction of a nation's economic growth. Exchange rates are therefore in turn 27 Therefore, exchange rate stability is one of the main factors influencing foreign investments, price stability and stable economic growth (Ajao, 2015).
In the recent years, Botswana's economy has been experiencing slow economic growth rate. From 2000 to 2018, the highest economic growth rate recorded was 11.3% in 2013, followed by 8.65% in 2010. In 2017 and 2018, the economy experienced low growth rates of 2.9% and 4.5% respectively (Statistics Botswana, 2018). Therefore, it is critical to recognise the impact of the volatile nature of the Rand on the Pula given the 45% weight it has in the currency basket. Economic question that arise is that, is the Rand/Pula exchange rate volatility responsible for crippling Botswana's rate of economic growth. Furthermore, Botswana is still a mineral-based economy and it has not succeeded in significantly diversifying its economy away from diamonds (Lewin, 2011). Therefore, it is important to evaluate, if the Rand/Pula exchange rate volatility has influence on Botswana's economic growth.
Exchange rate volatility can also have an indirect effect on the economic growth through its impact on the key determinants of the economic activity. These are trade flows, investment, and employment.
These effects may be, increases in transaction costs and decreases in international trade. It is assumed international trade and economic growth are positively associated. Volatile exchange rates make international trade and investment decisions more difficult because volatility increases exchange rate risk. Exchange rate risk refers to the potential to lose money because of a change in the exchange rate (Suranovic, 2006). Exchange rate volatility also decreases investments in a country, thus lowering the capital in a given economy (Tavlas, 2003). Hence, exchange rate volatility has a significant impact on economic growth. Therefore, it is crucial, to identify the impact of exchange volatility on a nation's economic growth. Hence this paper divulge the impact of the Rand/Pula exchange rate volatility on the economy of Botswana.
The paper proceeds as follows. Section II provides a brief background of Botswana's exchange rate system, volatile Rand and economic growth in Botswana. Section III is a review of relevant previous studies. Section IV provides the methodology. Empirical findings are discussed in Section V. Conclusions of the paper are given in Section VI.

Botswana's Exchange Rate System
Botswana adopted the crawling peg exchange rate mechanism in May 2005 (Bank of Botswana, 2016).The "crawling peg" exchange rate system, allows the foreign exchange rate to vary but only on the basis of a predetermined formula by sale and purchase of international reserves (McKenzie, 1983).
The use of a currency basket in place of a peg to a single currency, normally tends to stabilize a country's effective exchange rate.
The choice of an appropriate exchange rate regime is a paramount decision made to favour the progressive growth of any a nation's economy. The exchange rate policy issues have been a great area of concern in developing countries in recent years. This preceded the introduction of IMF and World 29 Bank stabilization and adjustment policies. It incorporated devaluation of exchange rate, introduction of new exchange rate management policies and trade liberalization measures (Atta et al., 1999).
The crawling peg exchange rate system allows the country to benefit from the advantages of the two extreme exchange rate regimes. Under the flexible exchange rate regime, the main advantage is invulnerability to currency crisis and the ability to absorb adverse shocks. On the fixed exchange rate system, the main advantage is that it promotes international trade and investment (Yagci, 2001).
Given a situation whereby the Pula had been allowed to float, large inflows of diamond revenues would have caused the Pula to appreciate. The appreciation of the Pula would have made non-mineral export sectors to be uncompetitive which would make economic diversification extremely difficult to achieve (Motlaleng, 2009). Furthermore, that would have triggered the economy to experience the undesirable Dutch disease. Dutch disease can arise due to an appreciation of the real exchange rate due to a natural resource boom from a tradable resource discovery. This reduces the international competitiveness of other tradable sectors as resource based exports crowd out commodity exports produced by those sectors. This may ultimately slow the growth of a country's exports. Consequently, it may harm a country's long-term economic growth goals (Barder, 2006). The ripple effects on the economy would be inconsistent with the nation's development and diversification objectives (Masalila & Motshidisi, 2003).
In an attempt to control the adverse effect of volatility of an independent float and the strait jacket of a fixed exchange rate, Botswana, has chosen an intermediate exchange rate regime. This enables her to enjoy the advantages of the two extreme exchange rate mechanisms. At the introduction of the Pula currency in 1976 Botswana had adopted a fixed but adjustable peg system. The Pula was pegged to the US dollar and, before 1980, the peg was revalued due to anti-inflationary reasons. The single currency peg coincided with a period in which the Rand was also pegged to the US dollar. Specifically, the exchange rate at which the Pula was pegged to the US dollar was also equal to that of the Rand against the Dollar. This demonstrated equality between the Pula and the Rand. This effect ended when the Rand was taken off the US dollar peg and allowed to float. To subdue the effects of exchange rate volatility between the Pula and the Rand, the Pula basket was introduced in 1980 (Masalila & Motshidisi, 2003). Given the volatile nature of the Rand, the Pula basket mitigated the volatility of the Rand by the less volatile currencies in the basket till the employment of the crawling peg exchange rate mechanism in May 2005. It is therefore apparent that the Pula has been facing the challenge of exchange rate volatility posed by the volatile Rand.

Volatility of the South African Rand
Volatility represents the degree to which a variable changes over time. The larger the magnitude of a variable change, or the more quickly it changes over time, the more volatile it is (Suranovic, 2006).
Exchange rate stability is one of the main factors influencing foreign investments, price stability and stable economic growth. The larger the fluctuations in an exchange rate are between two countries, the more volatile the exchange rate is described as (Bauwens et al., 2006). 30 www.scholink.org/ojs/index.php/ijafs International Journal of Accounting and Finance Studies Vol. 3, No. 2, 2020 The Rand has been a volatile currency (Miyajima, 2019;Fowkes et al., 2016;Hanusch et al., 2018).
South Africa's Rand has over the years, maintained the world's most volatile currency status. For the past few years, the South African Rand has seen a rapid depreciation (Oseifuah & Korkpoe, 2018).
Therefore this, will directly affect the Pula-Rand exchange rate.
Recent literature shows that the South African Rand is a volatile currency (for instance, Miyajima, 2019;Hanusch et al., 2018). Miyajima paper found that higher exchange rate volatility tends to increase core inflation but to a relatively limited extent in South Africa. He argues that the results support policy of allowing the rand to float freely. This is found to work as a shock absorber with South Africa's inflation targeting regime. The South African Rand has been relatively volatile due to domestic and external disturbances. The Rand acts as a shock absorber based on the South African Reserve Bank's free-floating exchange rate policy. For instance, domestic shocks have elevated the Rand volatility to above the US stock price volatility index. This is a commonly-used indicator of global uncertainty. The relatively high Rand volatility is argued to have an implication on South Africa's inflation (Miyajima, 2019 Therefore, this paper extends the frontiers of knowledge by investigating the impact of the Rand/Pula exchange rate volatility on Botswana's economic growth.

Botswana's Economic Growth
Economic growth is among the major key objectives of any economic policy. The other important variable in the formulation of economic policies is the exchange rate. This is because fluctuations of the real exchange rate can cause high fluctuation in the foreign trade and balance of payments.
Therefore, exchange rate systems, remains a key factor in the economic policy (Basirat et al., 2014). This was largely accounted for by the rapid growth of the diamond mining sector in the country. This was the period in which mining contributed the uppermost, recording a share of 50% to the GDP, in 1988/89 (Bank of Botswana, 2000). Thereafter the contribution of mining to the GDP declined leading to the overall decline in the GDP growth rate. The lowest growth rate the economy experienced was in 2009, at -7.7%. This was chiefly contributed to by the global economic recession. This was a ripple effect of the contraction of the mining sector, which contributed to only 24% in 2009, dwindling from 41.2% in 2008 (Bank of Botswana, 2009). One more significant plunge in the GDP growth rate was recorded in 2015, at -1.7%. This was also attributed to the decline in the mining output, which fell by 11.8% compared to the previous year. The 11.8% shrinkage was principally due to a 12.2% decline in the diamond production which was scaled down in response to the subdued global demand (Bank of Botswana, 2015). causality, meaning that exchange rate volatility affects economic growth and economic growth affects exchange rate volatility. Azid et al. (2005) investigated the impact of exchange rate volatility on growth and economic performance in Pakistan. The study used the GARCH to measure the volatility in the exchange rate series. The results of the study showed that excessive volatility of exchange rate regimes had no effect on the economic performance of manufacturing product in Pakistan. Sanginabadi and Heidari (2012) studied the effects of exchange rate volatility on economic growth of Iran. The study results showed a significant relationship between Iranian growth volume and real exchange rate volatility. The long-run results of ARDL model showed a negative effect of exchange rate volatility on economic growth. Oloyede and Fapetu (2018)

Generalized Method of Moments (GMM)
This paper employs the GMM to evaluate the impact of the exchange rate volatility on economic growth. GMM was first formalized by Hansen in 1982. The paper uses the Two Step System GMM by Arellano and Bover, (1995) and Blundell and Bond (1998). GMM is used to estimate this impact using the GARCH as a measure of exchange volatility. GMM is a method for constructing estimators, analogous to Maximum Likelihood (ML). GMM uses assumptions about specific moments of the random variables instead of assumptions about the entire distribution. This makes GMM more robust than ML, at the cost of some efficiency (Caner, 2009). The choice of GMM is also to overcome the 34 www.scholink.org/ojs/index.php/ijafs International Journal of Accounting and Finance Studies Vol. 3, No. 2, 2020 problem of endogeneity and simultaneity bias. Before employing the model, the properties of time series will be conducted using both Augmented Dickey fuller test (ADF) and the Philips-Perron test (PP) to determine the stationarity of the variables.

Empirical Results
This section presents the empirical results on the of impact exchange rate volatility on Botswana's economic growth and the Rand/Pula exchange rate volatility results. Section 5.1 presents the correlation matrix. Section 5.2 presents the unit root test results. The unit root test was used to determine the behaviour and characteristics of individual series of variables. The GARCH model results of exchange volatility given in Section 5.3. Lastly, Section 5.4 presents the empirical of the impact of the exchange rate volatility on Botswana's economic growth.

Correlation Matrix
Correlation matrix was used to determine the strength of relationship between the dependent variable (GDP) and the explanatory variables (independent variables). Table 1 shows the correlation between the variables used in this paper. The entries on the main diagonal give the correlation between one variable and itself while the entries off the main diagonal, give the pair-wise correlation among the variables. The pair-wise correlation is very low, this shows that there is no problem of multi-collinearity of the variables. The real exchange rate volatility variable (RER) shows that it is negatively related to GDP growth rate (GDP) as (R=-0.3532). Source: EViews 10 computations.

Unit Root Test Results
The process of testing for unit root in time series data is very important. A unit root test is conducted to check if a time series variable is non-stationary or not. In this paper, we begin with the Augmented  Vol. 3, No. 2, 2020 to under-reject the null hypothesis of no unit-root. Therefore, in confirmation to the outcomes of the ADF test, we also apply the Phillips-Perron (PP, 1988) unit-root test in (Gbatu et al., 2017. ** Significant at 5% significance level. *** Significant at 1% significance level.

LEVELS
The results of the unit root test using the ADF test in   ** Significant at 5% significance level.

The GARCH Model
The result of the GARCH test in Table 4   Note.

GMM Estimation Result
We employed the GMM model to examine the impact of the exchange rate volatility on Botswana's economic growth. The robustness of the results are reinforced with the diagnostic checks of the model in Table 5.
38  Labour Force (LF) is significantly related to economic growth rate. This shows that Labour Force is positively related to economic growth. These results are as per the expected sign and are consistent with the empirical results by Kargi (2014). The result show that, a percentage change in the Labour force will cause a 0.241% increase in economic growth. It must be noted that Labour Force and economic growth exhibit an inelastic relationship.
Trade Openness (OPEN) indicates a positive relationship with economic growth. A percentage change in the Trade Openness effects a 0.142% increase in the economic growth. Hence, Trade Openness and economic growth also exhibit an inelastic relationship. These results are as per the expected signs and consistent with the empirical results of Frankel and Romer (1999), Edwards (1998) and Sachs et al. (1995). This explains that trade openness positively contributes to the growth of Botswana's economy. Furthermore, results shows that inflation rate does not affect economic growth, ceteris paribus.
It is also shown that Foreign Exchange Reserves have an insignificant positive impact on economic growth. Lastly, Dummy Variable indicates that past negative economic growth rates adversely impact current growth rates. The R-square suggests that 69.93% of Botswana's economic growth is explained by the variation of the independent variables.

Conclusions
The results of the GARCH model proved the presence of volatility in the Rand/Pula exchange rate. The findings established the presence of time-varying conditional volatility of the Rand/Pula exchange rate.
The results further indicated the persistence of volatility shocks. The conclusive results from the GMM model suggest that the Rand/Pula exchange rate volatility has a negative but insignificant impact on Botswana's economic growth. This finding corroborates those of Kaur et al. (2019) and Musyoki et al. (2012). Their studies found negative but insignificant impact of exchange rate volatility on economic growth in Malaysia and Kenya, respectively. Our principal findings are that Botswana's economic growth is largely explained by trade openness and growth of labour force and not influenced by the Rand/Pula exchange rate volatility.