ICITI 2012

ISSN: 16941225

Exchange Rate Pass Through to Domestic Prices: Evidence from Mauritius by Verena TANDRAYEN-RAGOOBUR1 and Anjulee CHICOOREE Department of Economics and Statistics, University of Mauritius Abstract This study examines the impact of exchange rates and external shocks on import prices, domestic producer and consumer prices for the small island economy of Mauritius highly dependent on the external markets. Quarterly data from 1999 to 2010 and the Structural Vector Autoregressive model are used. The impulse response function suggests that exchange rate pass through to consumer prices is the highest; followed by producer prices while exchange rate pass through to import prices is the lowest. Our findings also reveal the existence of bidirectional causality only for the case of nominal effective exchange rate and producer prices. Further, the variance decomposition results indicate that the variance of import and producer prices is explained mainly by oil price shocks while the variance of consumer prices is largely accounted by import price shocks. This study shows that monetary policy has a determinant role in maintaining price level in the economy. Keywords: Exchange Rate Pass Through, Inflation, Structural Vector Autoregressive Model, Mauritius JEL Classification: F31, E31, C32

Address: Department of Economics and Statistics, Faculty of Social Studies and Humanities, University of Mauritius, Reduit, Mauritius. Phone Number: +230 7871282. Email: v.tandrayen@uom.ac.mu

ICITI 2012

ISSN: 16941225

1. Introduction Exchange rate pass through is the percentage change in local currency import prices resulting from a per cent change in the exchange rate between the exporting and importing countries (Goldberg and Knetter 1996). Exchange rate pass through can be either incomplete or complete and refers not only to the effect of exchange rate changes on import and export prices but also on consumer prices, producer prices, investments and trade volumes. The extent and degree of exchange rate pass through points out the importance of exchange rate fluctuations on domestic price inflation and also the extent to which exchange rates and import prices influence domestic inflation. Studies on exchange rate pass through investigate the impact of exchange rate fluctuations on domestic prices or more specifically on consumer price inflation (Choudhri and Hakura 2001, McCarthy 2006). McCarthy (2006) finds that exchange rate pass through to import prices is negative and particularly large for Belgium and Netherlands while pass through is small in Sweden and Switzerland. The exchange rate pass through to producer price index is fairly weak in most industrialized countries. The response of the consumer price inflation is the lowest except for Belgium and Netherlands. The variance decompositions results reveal that domestic policies are important in controlling domestic consumer inflation. Studies on emerging economies, namely Bhundia (2002) for South Africa shows that pass through elasticities are highest for import prices, followed by producer prices and consumer prices thus suggesting that exchange rate shocks have a successively smaller impact as one moves down the distribution chain. In addition shocks to producer prices have a considerable impact on consumer price inflation, so that policies oriented towards mitigating inflationary pressures at the producer price level may reduce consumer price inflation. Garcia and Restrepo (2001) show that the pass through to inflation is complete in the long run for Chile. Restrepo (2001) further concludes that exchange rate pass through depends positively on output gap and wages and that foreign price are positively related to inflation. Rowland (2003) indicates that exchange rate pass through is incomplete in Colombia even after a year where only 80 per cent of an exchange rate change is passed onto import prices and for producer and consumer prices exchange rate pass through is 28 per cent and 8 per cent respectively. Ghosh and Rajan (2007) study the evolution of exchange rate pass through into consumer price index for the Indian economy. The results indicate that exchange rate pass through elasticity of the rupeeUSD is between 45 per cent and 50 per cent and quite stable over the period under review. There is also evidence that exchange rate volatility consistently has a negative impact on exchange rate pass through. In addition, Sek and Kapsalyamova (2008) indicate that the degree of pass through is highest on import price, moderate on producer prices and lowest on consumer prices but on overall exchange rate pass through is incomplete in four East Asian countries namely, Singapore, Malaysia, Thailand and Korea before and after the financial crisis of 1997. The effect of import price shock is found to be stronger than an exchange rate shock in determining domestic prices especially in the case of Singapore. Like many developing countries, Mauritius depends on the rest of the world and the level of interdependence has increased in the last decade. Mauritius being a small island economy with a domestic market insufficiently large to support large scale production depends on imports from other countries to supply a large part of domestic consumption and on exports to other countries to provide markets for much of its output. It is highly vulnerable to any adverse economic changes in other economies. Mauritius has increasingly liberalised its trade frontiers leading to lower barriers to trade, for both goods and services. This has increased trade and intensified international competition. In addition to greater trade and financial liberalisation, two specific changes have impacted significantly the Mauritian economy namely the phasing out of the preferential access obtained on the EU market for sugar exports and the dismantling of the Multi Fibre Agreement for our textile products in 2005. In addition, currency markets show different degrees of volatility, reflecting the particular economic circumstances that the country faces through time. Exchange rate volatility is

ICITI 2012

ISSN: 16941225

another crucial element that needs to be considered for small countries that depend extensively on trade the case of Mauritius. Exchange rate changes have thus important implications on both producer and consumer price inflation. To our knowledge, there are no studies analysing exchange rate pass though for the small island economy of Mauritius which is highly dependent on trade and where food imports presently consists of 77 per cent of the total import bill. The main aim of the study is thus to investigate the extent and degree of exchange rate pass through to prices at different distribution levels from import prices, to producer prices up to consumer prices. The study further focuses on the existence and degree of causality between exchange rate and domestic prices and analyses the degree of exchange rate pass through to import, producer and consumer prices. Lastly, we assess the effect and importance of external shocks in explaining import and domestic prices in the Mauritian economy. A structural vector autoregressive (SVAR) model is estimated and the impulse response functions are used to calculate exchange rate pass through elasticity and also to analyse the effect of external shocks on import and domestic prices. Forecast error variance decomposition is equally obtained from the SVAR to examine the importance of external shocks in explaining import and domestic prices. The study is organised as follows. Section 2 reviews the theoretical and empirical literature on exchange rate pass through. Section 3 analyses the data used for Mauritius. Section 4 depicts the methodology adopted. Section 5 presents the econometric analysis and results. Section 6 concludes with policy recommendations.

ICITI 2012

ISSN: 16941225

2. Literature Survey The existing literature on exchange rate pass can be viewed from three perspectives. The first one studies exchange rate pass through into import prices for specific industries (Feinberg 1989 and Klitgaard 1999) while the second covers studies relating to exchange rate pass through into aggregate import prices (Hooper and Mann 1989; Campa and Goldberg 2005) and finally the last one regroups those studies that examine the pass through related to wholesale and consumer prices (McCarthy 2006). It is important to note that studies on exchange rate pass through at the macro level investigate exchange rate pass through from the monetary policy view where different prices indices (import price index, producer price index and consumer price index) are estimated. Our study captures all the three dimensions of exchange rate pass through for Mauritius. There are several possible channels through which exchange rate changes may affect prices. Changes in exchange rates have direct and indirect effects on consumer prices. At unchanged foreign currency costs of production, an exchange rate change affects the domestic currency price of foreign imports and this is termed as the direct effect. Changes in foreign import prices may affect consumer prices in two ways. The first effect is from the fact that imports may be finished goods used for consumption and thus the price level is directly influenced. Second, imports are also intermediate goods used in the production of domestic output. Thus, the appreciation of foreign currency increases domestic currency cost of production and this is in part passed through into higher output prices. The indirect effects of exchange rate fluctuations are divided into the competition effect and the wage inflation effect by Sachs (1985). The competition effect takes place when there are shifts in the demand for domestic output due to exchange rate fluctuations. An exchange rate appreciation leads to an increase in exports prices and fall in import prices. Given unchanged domestic costs, domestic producers have to react to lower import prices by cutting their own prices and profits margins due to lower competitors prices. On the other hand, exchange rate depreciation is likely to lead to an increase in the domestic demand for substitutes due to the rising import prices, thus putting upward pressure on prices of such products and causing consumer prices to rise. At the same time depreciation of the domestic currency make exports more competitive on the world market. The rising demand of exports leads to an upward pressure on the price of domestic tradable goods which also contribute to the rise in consumer prices. The last channel is the wage inflation effect. The wage inflation effect works through the determination of nominal wages which has a direct impact on production costs. Exchange rate depreciation increases the price of imported consumption goods leading to a fall in purchasing power of workers. To compensate for the fall in purchasing power, employers increase nominal wages and this leads to an increase in the cost of production which is consequently passed through into higher output prices. Laflche (1996) illustrates the direct and indirect effects of exchange rate fluctuations on the consumer price level in Figure 1 as follows.

ICITI 2012

ISSN: 16941225

Figure 1: Transmission mechanism of exchange rate depreciation to prices.

Exchange Rate Depreciation Direct effects Indirect effects

Imported inputs become more expensive

Imports of finished goods become more expensive

Domestic demand for substitutes rises

Demand for exports rises

Production costs rise Substitute goods and exports become more expensive

Demand for labour increases

Wage rises Consumer price rises

Source: Laflche (1996) The extent, speed and magnitude of exchange rate pass through to domestic prices depend on microeconomic and macroeconomic factors which may either deepen or lessen the impact of exchange rate volatility on domestic prices. One of microeconomic factors is the market structure (Dornbusch 1987; Menon 1996) as the responsiveness of prices to an exchange rate movement is determined by the relative number of foreign firms and the ratio of marginal cost to price of foreign suppliers. An appreciation of the domestic currency is found to lower price less than proportionally and the decline in the domestic price is larger the more competitive the industry and the larger the share of imports in total sales. Due to imperfect competition, pricing will no longer be at marginal cost and firms can charge a mark-up on costs to earn abnormal profits even in the long run. Pricing to market equally influences exchange rate pass through (Krugman 1986) as it is the pricing behaviour of firms exporting their products to a destination market following an exchange rate change. The extent of pricing to market is based on the competitive conditions in foreign markets. As such, exporters prefer profit margin to fluctuate rather than foreign currency prices in response to exchange rate fluctuations. The pricing strategies also influence exchange rate pass through. Producer currency pricing implies that in the short run consumers prices change one-for-one with changes in the nominal exchange rate and thus the law of one price holds while in the case where nominal prices are set in advance in the currency of the consumer, this is referred to as local currency pricing where nominal exchange rate changes do not lead to a change in prices in the short run. Therefore, in the producer currency pricing case there will be complete exchange rate pass through while in the local currency pricing case exchange rate pass through will be zero in the short run. One of the macroeconomic factors influencing exchange rate pass through is the inflationary environment of a country. Countries with high inflation tend to have more persistent costs compared to countries with low inflation. Thus, a high inflationary environment would tend to increase the exchange rate pass through to domestic prices. A positive and significant association between pass through and the average inflation rate may hold across countries (Choudhri and Hakura 2001). Further, firms usually respond more to cost increases if cost changes are perceived to be more persistent (Taylor 2000). Further, exchange rate volatility 4

ICITI 2012

ISSN: 16941225

also influences exchange rate pass through (Mann 1986). Due to foreign exchange uncertainty foreign firms cannot perfectly forecast exchange rate so their profit margins are exposed to exchange rate shocks. For instance, if exporters perceive a shock to be transitory, they would avoid changing prices by shifting the markup thus reducing the pass through. While if firms expect exchange rate shocks to be persistent, they are more likely to change prices rather than adjust profit margins. In line with the inflationary environment, monetary policy environment influences exchange rate pass through to domestic prices. From Devereux et al. (2004) model of endogenous exchange rate pass through, countries with successful policy of monetary stabilisation are able to reduce the variance of their money growth and thus have low rates of exchange rate pass through. This is explained by foreign exporters who prefer to set their prices in that countrys currency leading to the producer currency pricing case, thereby reducing the impact of exchange rate changes on the countrys consumer price index. Low variability of monetary shocks reduces the predictive power of exchange rate in determining monetary shocks and this effect therefore suggests another reason for the pass through to be smaller under a low inflation environment (An 2006). McCarthy (1999) indentifies openness as impacting on exchange rate pass through. The greater the degree of openness, the larger the pass through (Soto and Selaive 2003). Aggregate demand uncertainty where foreign firms cannot perfectly forecast gross domestic product or gross national product is another economic variable that affects exchange rate pass through (Mann 1986). The profit margins of foreign firms are exposed to demand shocks. Exporters usually alter the profit margins when aggregate demand shifts due to exchange rate fluctuations in an imperfectly competitive environment, thus reducing pass through. Thus pass through is likely to be lower in countries where aggregate demand is more volatile. 3. Data Source and Analysis 3.1. Data Source Data from both the Statistics Mauritius and the Bank of Mauritius are used. Oil price, money supply and short term interest rate are obtained from the Bank of Mauritiuss various annual reports and monthly statistics bulletins. Gross domestic product, import price indices, producer price indices and consumer price indices are obtained from the Statistics Mauritius. The nominal effective exchange rate is calculated using data on imports, exports and exchange rate from the Bank of Mauritiuss various annual reports. 3.2. Data Analysis The nominal effective exchange rate volatility of Mauritius is the standard deviation of the nominal effective exchange rate characterised by periodical fluctuations where at first exchange rate volatility increased at a diminishing rate and from 2003-04 onwards the nominal effective exchange rate volatility experienced large swings as illustrated in figure 2 below. The increase in volatility in 2004-05 reflects the local market conditions and international trends where the rupee depreciated against the currencies of major trading partners while the year 2005-06 was subject to a fall in exchange rate volatility. Volatility peaked in the year 2006-07 and it stood at 0.21 which reflected local market conditions such as the increase in the Lombard rate, shortage of liquidity on the domestic interbank foreign exchange market, intervention of the Bank of Mauritius in the foreign exchange market and international trends. From 2006-07 to 2008-09 exchange rate volatility had a downward trend but pick up again in 2009-10 to reach 0.11 and it was due to the central banks non intervention in the market to reduce volatility of the exchange rate and also because of the lack of liquidity on the spot foreign exchange market. 5

ICITI 2012

ISSN: 16941225

Figure 2: Yearly nominal effective effective exchange rate volatility over the period 1999 to 2010
Yearly SD of NEER 0.25 0.2 0.15 0.1 0.05 0

SD

199900

200001

200102

200203

200304

200405

200506

200607

200708

200809

Source: Authors Computation The imports behaviour, consumer and producer price behaviour over the period 1999 to 2010 in figure 3 below shows that imports have been increasing and peaked in the year 2007-08 while during the last two fiscal years imports have been declining and the consumer price index2 and producer price index had a increasing trend over the years. Figure 3: Import expenditure, consumer and prodcuer price index over the period 1999-2010

Rs millions
150000 120000 90000

Imports (Rs Million)

CPI

PPI

200910

Year

Index
160.0

Right scale

Left scale
120.0 80.0

60000 30000 0

Right scale
40.0 0.0

99-00 00-01 01-02 02-03 03-04 04-05 05-06 06-07 07-08 08-09 09-10

Year

Source: Data Bank of Mauritius Annual reports, CSO, Authors computation

There have been considerable changes in the composition of imports over the years as illustrated in table 1. Most of the ten categories of imports have been rising and the most noted rise in food and live animals is found where it increased from 12.43 per cent in 1999-00 to 17.76 per cent in 2009-10. However, the proportion of manufactured goods classified chiefly by materials and machinery and transport equipment in total imports decreased over the studied period.

The CPI is of base year 2006-07 while the PPI is of base year 2007

ICITI 2012

ISSN: 16941225

Table 1: Import Composition from 1999 to 2010 Financial Year (per cent) 99-00 02-03 05-06 09-10 12.43 16.45 14.73 17.76 0.83 0.85 0.89 1.62 2.91 2.51 2.41 2.69 9.55 10.93 17.84 17.61 0.85 1.15 0.73 1.02 7.46 8.19 7.62 8.74 31.57 28.34 19.77 17.89

Food and Live Animals Beverages and Tobacco Crude Materials, Inedible except Fuels Mineral Fuels, Lubricants and Related Materials Animal and Vegetable Oils and Fats Chemicals and Related Products Manufactured Goods classified chiefly by Materials Machinery and Transport Equipment 25.12 21.78 28.08 22.85 Miscellaneous Manufactured Articles 9.07 9.64 7.59 9.29 Commodities and Transactions not classified 0.22 0.18 0.34 0.54 elsewhere in the SITC Source: Data Bank of Mauritius Annual reports, Authors Computation Imports consist of consumer goods and raw materials used in the manufacturing industry, changes in imports expenditures on these products may have an impact on the consumer price index and producer price index (manufacturing). Imports expenditure have been increasing over the years where the highest increase of 21 per cent was recorded in the year 2004-05 and it was also noted that consumer and producer price inflation was higher at 5.58 per cent and 7.85 per cent respectively. High energy prices and the depreciation of the rupee are seen to have led to the rise in imports expenditures and domestic inflation. The highest rate of consumer inflation was registered in 2006-07 led by the depreciation of the rupee and hikes in food prices where imports expenditures and producer price inflation equally rose. Producer price inflation was highest in the year 2007-08 due to high import. 2007-08 saw a stabilising in the inflation rate as well as a fall in imports expenditures while in the last fiscal year import expenditures and consumer price inflation were subject to a rise while producer price inflation fell. 4. Methodology The objectives of the study are first to examine the degree of the pass-through of exchange rates in Mauritius to prices of goods at various distribution levels, from the prices of imported goods to producer prices and to consumer prices. Second, we analyse the impact and extent of external shocks namely oil prices shocks, exchange rate shocks and import price shocks on consumer and producer price. Lastly, we investigate the impact and extent of oil price shocks and exchange rate shocks on import prices. 4.1. Model Specification Our model derives from McCarthy (2006) which is a model of pricing along distribution chain - import, production and consumer - to track pass through from exchange rate fluctuations at each stage of the distribution chain. The model also considers supply, demand and exchange rate shocks. Furthermore, the model accounts for the reaction of the Central Bank to carry out monetary policy. The supply shocks are identified from the dynamics of oil price inflation denominated in US dollars ( . The proxy for demand shocks is the dynamics of output gap ( ) in the country after accounting for the contemporaneous effect of the supply shock. Finally the 7

ICITI 2012

ISSN: 16941225

exchange rate shocks ( ) are identified from the dynamics of exchange rate changes ( after taking into account the contemporaneous effect of demand and supply shocks.

Inflation at each stage import, producer and consumer at time period t, is assumed to contain several components where the first component is the expected inflation at that stage based on available information at the end of the time period t-1. The next two components are the effect of period t domestic supply and demand shocks on inflation at that stage. The fourth component shows the effects of exchange rate shocks on inflation at that stage. Then the effects of the shocks of the previous stages of the chain are included and finally there is the stages shock. The shocks at each stage can be considered as the changes in the pricing power and markups of firms at these stages. Finally, the Central Banks reaction function is estimated where the short term interest rates ( ) are related to the previously mentioned variables in the model and the money demand function relates money growth ( to the other variables in the model. The model is as follows: ) + = ) + + + + (1) (2) (3) (4) (5) (6) (7) (8)

where and are import price, producer price and consumer price inflation respectively and , , are import price, producer price and consumer price shock respectively. is the monetary policy shocks and is the money demand shock. is the expectation of the variable based on the information set available at end of period t-1. 4.2. Variables The variables used for the study spans from the first quarter of 1999 to the second quarter of 2010. Oil prices are obtained from the Bank of Mauritius and are the quarterly average of IPE3 Brent Oil in US dollars per barrel. The output gap is calculated as the deviation of the log of real gross domestic product from a quadratic trend. The nominal effective exchange rate is calculated as follows:

where NEERt is the quarterly nominal effective exchange rate at time t.

IPE is the international petroleum exchange in London which is now known as ICE, the Intercontinental Exchange in London.

ICITI 2012

ISSN: 16941225

is the quarterly weight of trading partner i in total trade at time t. The weight of the first ten trading partner which have the highest share in total trade is taken for the calculation of the nominal effective exchange rate. The ten trading partners on average account for 70 per cent of total trade of Mauritius. is the end of quarter exchange rate of trading partner i at time t. The weights of trading partners are calculated using the following formula: =1 + )

where is the quarterly import of the Mauritius from country i at time t, is the quarterly export of Mauritius to country i at time t, + ) is quarterly total export plus imports from all trading partners at time t. The import price inflation is calculated from the quarterly average of IPI on a base year 2007, the producer price inflation is calculated from the quarterly average of the producer price index with the base year of 2007 while the consumer price index inflation is calculated from the quarterly average of consumer price index of base year 2006-07. From Frimpong and Adam (2010), the quarterly average of the weighted average yield of 91 days t-bills is used as the short term interest rate to reflect the central banks behaviour. The money supply variables used are the quarterly average M2 from the first quarter of 1999 to the second quarter of 2006 and from the third quarter of 2006 onwards the Broad Money Liabilities are used4. 4.3. Estimation Technique To estimate equation (1)-(8), it is assumed that the conditional expectations in these equations can be replaced by linear projections on lags of the eight variables in the system. Therefore, the model can be expressed and estimated as a VAR using a Cholesky decomposition to identify the shocks. The structural VAR model is represented as follows: (9) where is the N x 1 vector of contemporaneous endogenous variables (oil price, output gap, nominal effective exchange rate, price indices, short term interest rates and money supply variable), the matrix is of order N x N and describes the contemporaneous relationships between the variables. A(L) is the lag polynomial matrix of order infinity, is the (unobserved) vector of structural shocks of order N x 1. By multiplying equation (4.9) by an inverse matrix , we obtain the reduced form of the VAR model. It must be noted that this adjustment is necessary because the model given in equation (4.9) is not directly observable and structural shocks cannot be correctly identified. Therefore, (10)

The Bank of Mauritius stopped the publication of M2 as from third quarter of 2006. From the third quarter of 2006 till date Bank of Mauritius reports broad money liabilities.

ICITI 2012

ISSN: 16941225

where is a N x 1 vector of serially uncorrelated structural disturbances (errors) of the model and it is obtained as follows: or (11)

n(n-1)/2 restrictions are imposed on the matrix A0 based on the Cholesky decomposition of the residual variance covariance matrix which defines A as a lower triangular matrix. The lower triangular matrix implies a recursive scheme among variables and also that some structural shocks have no contemporaneous effect on some endogenous variables given the ordering of the endogenous variables. Following the identification scheme, equation (11) can be written as:

0 0

0 0 0

0 0 0 0

0 0 0 0 0

0 0 0 0 0 0

0 0 0 0 0 0 0 (12)

where

are the reduced form VAR residuals.

It is important to determine the order of the variables so as to identify structural shocks. As shown by equation 12, oil prices are ordered first because the reduced form residuals of oil prices are not likely to be affected contemporaneously by any other shocks except oil price shock itself while oil price shocks affect all the variables in the system contemporaneously. Output gap is ordered next, then the nominal effective exchange rate. The import price index in ordered next as the exchange rate is at first going to have an effect on import prices. Below the import price index is the producer price index because the import prices will have an effect on producer prices as producers import inputs from abroad. The next price index is consumer price where it is directly affected by import prices and also indirectly affected by producer prices. The last two variables which are short term interest rate and growth in money supply is ordered next to illustrate the reaction of the central bank following inflation and also to maintain the value of the rupee against other currencies. The exchange rate pass through to domestic prices is calculated from the impulse response function results and the pass through elasticity at t is given by: Pass through elasticity at time t

= per cent change in the price level t quarters after the shock (13)
Initial per cent change in the exchange rate at time t=0 10

ICITI 2012

ISSN: 16941225

The numerator is the percentage change in the level of the respective price indices between the period zero, when the initial exchange rate shock hits, at time t. The denominator is the percentage change in the nominal effective exchange rate at time 0. 4.5. Expected Results From the work of Bhundia (2002), Faruqee (2006) and Sek and Kapsalyamova (2008) among others, it can be expected that first exchange rate pass through to import prices is higher compared to exchange rate pass through to producer and consumer prices and second exchange rate pass through to producer prices is higher compared to exchange rate pass through to consumer prices. Thus the effect of an exchange rate shock declines along the distribution chain. Studies carried out by Bhundia (2002), Garcia and Restrepo (2001) and Rowland (2003) in developing countries indicate low exchange rate pass through in domestic prices. Although exchange rate pass through is different across countries but the main similarity is that exchange rate pass through to consumer prices is the lowest compared to exchange rate pass through in producer and import prices. Such evidence has also been found in emerging countries like Brazil where Belaisch (2003) results indicate that exchange rate pass through to consumer price is 17 per cent after a year while around 100 per cent for wholesale prices. The effect of external shocks in explaining consumer prices varies across countries where for the case of industrialised countries, McCarthy (2006) notes that external factors explain a fairly small proportion of the variance of domestic consumer prices while Bhundia (2002) shows that shocks to producer prices tend to have a considerable impact on consumer prices for the case of South Africa. In four of the East Asian economies selected by Sek and Kapsalyamova (2008), it was observed that that import price shocks have a stronger effect than exchange rate shocks in determining domestic prices. 5. Findings 5.1. Unit Root Tests, Diagnostic Tests and Granger Causality Tests The unit root tests indicate that all variables are stationary except the short term interest rate variable which is in our case t-bills yield is found to be integrated of order one. Given the autoregressive nature of the model, the lag order is determined using the maximum likelihood ratio and also such that there is no autocorrelation among residuals. The likelihood ratio results indicate the inclusion of 3 lags. To determine whether there is a cointegrating relationship among the variables the Johansen (1995) test is applied. Although evidence of cointegration for rank 3 is found, it is ignored as no long run relationship among the variables is studied. The diagnostic tests on the structural VAR indicate no autocorrelations among residuals, residuals are normally distributed and the model is stable as the characteristic roots have a modulus less than one and all eigenvalues lie inside the unit circle. The granger causality test undertaken to examine the existence of causality from exchange rates to imports and domestic prices and vice versa and also from imports prices to domestic prices and vice versa indicates that there is bidirectional causality in one case where producer prices granger cause nominal effective exchange rate and vice versa while it is found none of the variables; nominal effective exchange rate, import and producer prices granger cause consumer prices.

11

ICITI 2012

ISSN: 16941225

5.2. The Effect of Exchange Rate Shocks on Import and Domestic Prices Figure 4: Exchange Rate Shocks Impacts on Import, Producer and Consumer Prices Import Prices
erpt01, neer, ipi
2

-2

-4 0 5 10 15

step 95% CI cumulative orthogonalized irf

Graphs by irfname, impulse variable, and response variable

Producer Prices
erpt01, neer, ppi
2

-2

-4 0 5 10 15

step 95% CI cumulative orthogonalized irf

Graphs by irfname, impulse variable, and response variable

12

ICITI 2012

ISSN: 16941225

Consumer Prices
erpt01, neer, cpi
2

-2

-4

10

15

step 95% CI cumulative orthogonalized irf

Graphs by irfname, impulse variable, and response variable

An exchange rate shock at time zero of 9.48 per cent (calculated from the change in exchange rate equation) leads to a fall in the nominal effective exchange rate, indicating an appreciation of the exchange rate. It is thus expected that the negative exchange rate shock will lead to a fall in import, producer and consumer price inflation. The cumulative orthogonalised impulse responses show the impact of an exchange rate shock at time zero on the different prices and are illustrated in figure 4. An exchange rate shock leads to a fall (as expected) in import prices, although the impact on import prices in not immediately felt. At time zero the effect of the exchange rate shock on import price is 0.64 per cent, in three quarters it is -0.19 per cent and the cumulative effect increases further to -0.29 per cent in six quarters and by the end of the 12 quarters, the exchange rate shock leads to a decrease in import price of -1.04 per cent. The exchange rate shock has the expected impact on producer prices where the instant impact leads to a fall in producer prices of -0.11 per cent and increases to -0.32 per cent and -0.77 per cent by the end of three and six quarters respectively. The cumulative impact increases to -1.45 per cent by the nine quarters while over the 12 quarters the effect of the exchange rate shock increases to -1.34 per cent. The exchange rate shock has a negative impact on consumer prices where at time zero consumer prices fall by -0.48 per cent and maintain its downward trend. In the case of consumer prices the effect of an exchange rate shock is much higher compared to import and producer prices where by six quarters consumer prices have fallen by -1.19 per cent and over three years the cumulative impact on consumer prices is -2.26 per cent.

13

ICITI 2012

ISSN: 16941225

5.3. Exchange Rate Pass Through Elasticity to Import and Domestic Prices From figure 5 below, exchange rate pass through is incomplete in the three price indices. However, the behaviour of the indices differs. In the case of import prices, the initial pass through elasticity is negative, implying that there is no immediate impact of an exchange rate change on import prices. Over 3 quarters, the pass through increases to around 1.97 per cent and to 3.04 per cent in 6 quarters following the initial exchange rate shock. By the end of 12 quarters the exchange rate pass through to import prices is 10.93 per cent. Thus in the short term (3 quarters) exchange rate pass through is low, while in the long term (over 3 years) exchange rate pass through increases slightly. It can be observed that exchange rate shocks have little impact on import prices both in the short and long term. Figure 5: Exchange Rate Pass Through Elasticity to Import and Domestic Prices
30.00 25.00 20.00 15.00 10.00 5.00 0.00 -5.00 -10.00

0 -6.74 1.14 5.06

3 1.97 3.40 8.97

6 3.04 8.16 12.56

9 4.41 15.28 18.67

12 10.93 14.13 23.87

IPI PT Elasticity PPI PT Elasticity CPI PT Elasticity

Quarter

To better understand the reasons behind the low exchange rate pass through to import prices, it is important to know the sources of Mauritian imports. Mauritius imports are from three main countries namely; the European Union, India and China and these countries accounts respectively for 21.68 per cent, 18.72 per cent and 12.60 per cent of the total imported goods for the year 2009. Another important aspect is that the major imports in 2009 were machinery and transport equipment, food and live animals, manufactured goods classified chiefly by material and mineral fuels, lubricants and related products. Most of the imported machinery and transport equipment are from the European Union, China and Japan. A large proportion of imported food and live animals are from Europe and Australia. The sources of main imports are India and China where price is expected to be low given the cheap labour available in these countries. In addition, low quality products are imported from China and this further explains the low import prices. Exchange rate pass through to producer prices is positive at time zero and increases to 3.40 per cent over 3 quarters following the initial exchange rate shock and over 9 quarters exchange rate pass through to producer prices amounts to 15.28 per cent, which is the highest pass through over three years. Over 12 quarters exchange rate pass through to producer prices fall slightly to reach 14.12 per cent. Exchange rate shocks have an important impact on producer prices where in the short term exchange rate pass through is 3.40 per cent while in the longer term the pass through is 14.13 per cent. The exchange rate pass through to producer prices is explained by the fact that imports consist of machinery and transport equipment, manufactured goods classified chiefly by materials and mineral fuels, lubricants and related products which are mainly used for production by domestic manufacturing firms. Thus exchange rate changes are likely to have an impact on the price of these imported goods which leads to an increase in the producer price. 14

ICITI 2012

ISSN: 16941225

Exchange rate pass through to consumer prices is considerably higher compared to the other two price indices. An exchange rate change at time zero leads to an increase in consumer prices by 5.06 per cent. Pass through in the 3 and 6 quarters after the initial shock is 8.97 per cent and 12.56 per cent respectively. The pass through effect to consumer prices is 23.87 per cent after three years. Consumer prices are thus more impacted by exchange rate shocks where the short term exchange rate pass through is 8.97 per cent and 23.87 per cent in the long term. Therefore, exchange rate shocks have the largest impact on consumer prices followed by producer prices and the lowest impact is felt on import prices. Changes in exchange rate are found to have the largest impact on consumer prices. The fact that Mauritius imports most of the food products, exchange rate changes have important implications on the price of foodstuffs. A higher exchange rate pass through to consumer prices can equally be explained by the change is exchange rate regime from managed float to a flexible system where the Central Bank intervenes in the market only to smooth out fluctuations rather than to maintain the value of the rupee against other currencies within a range. Thus exchange rate changes have an impact on consumer prices. Mauritius being a resource scarce economy, consumer goods are imported and thus despite rising prices, local consumers will still buy imported consumer goods which are not produced domestically. Thus, an exchange rate depreciation which increases import prices will cause importers to pass the high import prices to consumers in the form of higher prices. 5.4. The Effect of External Shocks on Import and Domestic Prices The Mauritian economy is highly vulnerable to external shocks. Besides exchange rate shocks which affect import and domestic prices, other external shocks such as supply shocks and import price shocks affect domestic prices. Impact of External Shocks on Import Prices For the case of import price, the effect of supply shock (proxied by oil price inflation) on import price is given by figure 6 below Figure 6: Impact of Oil Price Shock on Import Prices
erpt01, oil, ipi
6

10

15

step 95% CI cumulative orthogonalized irf

Graphs by irfname, impulse variable, and response variable

The initial effect of an oil price shock leads to an increase in import prices of around 0.67 per cent and increased further to 2.13 per cent 3 quarters after the initial shock. The impact of the shock decelerated in the next 5 quarters while the impact increased as from 9 quarters after the initial shock to reach 2.86 per cent in three years. Thus supply shocks seem to affect import prices both in the short and long term.

15

ICITI 2012

ISSN: 16941225

Impact of External Shocks on Producer Prices Figure 7 below shows the impact of oil price shocks and import price shocks on producer prices. The positive oil price shock has a negative effect of 0.10 per cent on producer prices at time zero implying that the impact of an oil price shock is not immediately felt on producer prices. The cumulative impact increases to 0.70 per cent and 1.16 per cent 3 and 6 quarters respectively after the initial shock and after 3 years producer price increases by 2.04 per cent following the supply shock at time zero. A negative import price shock at time zero leads to a fall in producer prices by -0.2 3per cent. The impact increases to -0.55 per cent after 6 quarters and finally after 3 years the impact of the import price shock leads to a fall in producer prices by -1.14 per cent. Figure 7: Impact of Oil Price Shock and Import Price Shocks on Producer Prices Oil Price Shocks
erpt01, oil, ppi
6

10

15

step 95% CI cumulative orthogonalized irf

Graphs by irfname, impulse variable, and response variable

Import Price shocks

erpt01, ipi, ppi
2

-2

-4 0 5 10 15

step 95% CI cumulative orthogonalized irf

Graphs by irfname, impulse variable, and response variable

Impact of External Shocks on Consumer Prices The impact of external shocks on consumer prices is illustrated by figure 8 below. Following a positive oil price at time zero consumer prices increase. Although a negative effect is felt on consumer price from 1 to 3 quarters after the initial shock, the cumulative effect of the oil price shock after three years leads to an increase in consumer prices by around 1.61 per cent. A negative import price shock at time zero initially leads to an increase in consumer prices while the desired impact is felt as from two quarters after the initial shock. Over 6 quarters the cumulative impact of the import price shock is 0.14 per cent and over 12 quarters consumer prices have fallen by -0.03 per cent following an import price shock at time zero.

16

ICITI 2012

ISSN: 16941225

Figure 8: Impact of Oil Price Shock and Import Price Shock on Consumer Prices Oil Price Shocks
erpt01, oil, cpi
4

-2 0 5 10 15

step 95% CI cumulative orthogonalized irf

Graphs by irfname, impulse variable, and response variable

Import Price Shocks

erpt01, ipi, cpi
4

-2 0 5 10 15

step 95% CI cumulative orthogonalized irf

Graphs by irfname, impulse variable, and response variable

Oil price shocks have a larger impact on import prices followed by producer prices and then consumer prices. In the case of import prices, an import price shock has a larger influence on producer rather than consumer prices. This analysis therefore puts forward that domestic prices are vulnerable to external shocks and this can be explained by the high degree of openness of Mauritius. The impulse response functions indicate the extent of pass through of external shocks to import and domestic prices but do not indicate how important these shocks are in explaining import and domestic prices. Thus the forecast error variance decomposition (FEVD) is used to analyse the importance of externals shocks in explaining fluctuations in import and domestic prices.

17

ICITI 2012

ISSN: 16941225

5.5. Forecast Error Variance Decomposition of Import and Domestic Prices From figure 9 below, oil price shocks are important in explaining import price variance where in the second quarter oil price shocks explain 25.45 per cent of the variance in import prices. In the following quarters the importance of oil price shocks in explaining the variance in import prices fell and reach 15.07 per cent in three years. Nominal effective exchange rate explains a smaller proportion of the variance of import prices. In the first quarter 6.31 per cent of the variance in import prices is explained by exchange rate shock. By 6 quarters the proportion increased to 8.65 per cent while by the end of three years the variance of import price explained by exchange rate shock increased to 10.01 per cent. Oil price shock is more important in explaining the variance in producer prices followed by exchange rate and import price shocks. In the first quarter oil prices explain 0.63 per cent of the forecast variance and as the forecast horizon increases the percentage increases to 9.89 per cent in 6 quarters and over three years to 11.99 per cent. Initially exchange rate explained a fairly small proportion of the variance of producer price around 0.70per cent but as the forecast period increased the per cent increases to 8.33 per cent in 9 quarters and reached 11.62 per cent in three years. Import prices explain very little of the variance of producer prices where initial import prices explains 3.33 per cent of forecast variance and over three year the percentage increased slightly to reach 5.97 per cent. The variance of consumer prices is mostly explained by import prices followed by oil prices and exchange rate. Initially import price explains 7.84 per cent of the variance in consumer prices and this percentage increases over three years to reach 16.02 per cent. Oil prices explain a smaller proportion of the variance in import prices which stand at 0.53 per cent initially and increases to 8.67 per cent by three years. The smallest proportion of the variance in consumer price is explained by exchange rate. Although at first exchange rate explained a larger proportion of the variance of consumer price compared to oil price, but over the forecast horizon oil price overtook exchange rate where over three years exchange rate explains 4.99 per cent of the variance in consumer prices.

18

ICITI 2012

ISSN: 16941225

Figure 9: Forecast Error Variance Decomposition Results

Percentange of IPI forecast variance attibuted to external shocks

30 25 20 % 15 10 5 0 1 2 3 4 5 6 7 OIL 8 9 10 NEER 11 12 Forecast Horizon

Percentange of PPI forecast variance attibuted to external shocks

16 14 12 10 8 6 4 2 0 1 2 3 4 5 OIL 6 7 8 9 NEER 10 11 12 IPI

Forecast Horizon

Percentange of CPI forecast variance attibuted to external shocks

18 16 14 12 10 8 6 4 2 0 1 2 3 4 5 6 OIL 7 8 9 NEER 10 11 12 IPI

Forecast Horizon

19

ICITI 2012

ISSN: 16941225

6. Conclusion This study examines the degree and extent of exchange rate pas through to import and domestic prices and also the effect of external shocks such as oil price shocks and import price shocks on domestic prices. So as to understand which shock better explains the variance in import and domestic prices, the forecast error variance decomposition of each price index have equally been studied. Another investigation carried out in this study is on the existence and degree of causality from exchange rate to domestic and import prices and vice versa. Using a structural VAR model that incorporates a distribution chain, the results from granger causality test indicate that bidirectional causality exist is one case where producer price granger causes nominal effective exchange rate and vice versa while in other cases unidirectional causality is found. It is equally found that nominal effective exchange rate, import and producer prices do not granger cause consumer prices. The impulse response functions of the structural VAR are used to calculate exchange rate pas through elasticity and the results indicate that exchange rate pas through to consumer prices is highest and not complete. The second larger impact of an exchange rate shock is felt on producer prices and the smallest on import prices. Therefore pass through increases as one goes along the distribution chain. The findings especially for the case of consumer and import prices are not in line with other studies. But this can be explained by the fact that main sources of imports being China and India, lower import prices can be expected given the cheap labour and the low quality products from China. Moreover, the pricing to market practice can explain the low pass through to import prices. As far as consumer prices are concerned the reason behind the high pass through is the dependence of households consumption on imported consumer goods where most of them are not subsidised. The increasing cost of living in Mauritius and the change in the exchange rate regime can also explain the high pass through in consumer prices. Externals shocks equally have an effect on import and domestic prices. The results suggest that oil price shocks have a larger impact on import prices compared to producer and consumer prices while import price shocks have a larger influence on producer prices. The examination of external shocks in explaining the variance in the price indices pointed out that the variance of import and producer prices are explained mainly by oil price shock while for the case of consumer prices, import prices explain most of its variance. External shocks thus play an important role in the Mauritian economy whereby prices are exposed to external shocks. However, Mauritius being a small island developing state is vulnerable to external shocks and thus cannot do much to reduce the impact of these external shocks on the economy given the extent of openness of the country. The study can be useful for policy makers such as Central Bank, in controlling the price level in Mauritius. The indication that exchange rate pass through to consumer prices is highest but still incomplete implies that domestic policies such as monetary policy have a significant role in controlling consumer price inflation. Given that import price shocks have an important impact on producer prices, managing inflation at the level of imports could be effective to reduce producer price inflation. Moreover, the variance of consumer prices is explained mainly by import prices. Thus, managing inflation at import level and the monitoring of pricing technique by importers can contribute in reducing consumer price inflation.

20

ICITI 2012

ISSN: 16941225

References An, L., 2006. An,Exchange Rate Pass-Through:Evidence Based on Vector Autoregression with Sign Restrictions. MPRA Paper No. 527. Belaisch, A., 2003. Exchange rate pass-through in Brazil. IMF Working Paper, WP/03/141. Bhundia, A., 2002. An Empirical Investigation of Exchange Rate Pass-Through in South Africa. IMF Working Paper African Department, WP/ 02/165. Campa, M.J. and Goldberg, L.S., 2005. Exchange rate pass-through to import prices in the Euro area. Federal Reserve Bank of New York Staff Reports 219. Choudhri, E.U. and Hakura, D.S., 2001. Exchange Rate Pass through to Domestic Prices: Does the inflationary environment Matter? IMF Working Paper WP/01/194. Devereux M.B., Engel, C. And Storgaard, P.E., 2004. Endogenous exchange rate passthrough when nominal prices are set in advance. Journal of International Economics, 63 (2004), 263 291. Dornbusch, D., 1987. Exchange Rate and Prices. The American Review, 77 (1), 93-106. Faruqee, H., 2006. Exchange rate pass-through in the Euro Area. IMF Staff Papers, 53 (1). Feinberg, R.M., 1989. The Effects of Foreign Exchange Movements on U.S. Domestic Prices. Review of Economics and Statistics 71 (3), 505-11. Garcia, C.J. and Restrepo, J.E., 2001. Price Inflation and Exchange Rate Pass through in Chile. Central Bank of Chile Working Papers, No 128. Ghosh, A. and Rajan, R.S., 2007. Macroeconomic Determinants of Exchange Rate PassThrough in India. Colorado College Working Paper 2007-02. Goldberg, P.K. and Knetter, M.M., 1996. Goods Prices and Exchange Rates: What have we learned? NBER Working Paper Series No. 5862. Hooper, P. and Mann, L.C., 1989. Exchange Rate Pass-Through in the 1980s: the Case of U.S. Imports of Manufactures. Brookings Papers on Economic Activity, 20 (1989-1), 297337. Klitgaard, T., 1999. Exchange Rates and Profit Margins: The Case of Japanese Exporters. Federal Reserve Bank of New York Economic Policy Review, 5(1), 41-54. Krugman, P., 1986. Pricing to market when the exchange rate changes. NBER Working Paper Series No. 1926. Laflche, T., 1996. The impact of exchange rate movements on consumer prices. Bank of Canada Review, Winter 1996-1997, 21-32. Mann, C.L., 1986. Prices, Profits Margins, and Exchange Rates. Federal Reserve Bulletin, 72 (6), 366-79. McCarthy, J., 2006. Pass through of exchange rates and import prices to domestic inflation in some industrialized economies. Federal Reserve Bank of New York. 21

ICITI 2012

ISSN: 16941225

McCarthy, J., 1999. Pass through of exchange rates and import prices to domestic inflation in some industrialized economies. BIS Working Paper, No.79. Menon, J., 1996. The Degree and Determinants of Exchange Rate Pass-Through: Market Structure, Non-Tariff Barriers and Multinational Corporations. The Economic Journal, 106 (435), 434-444. Rowland, P., 2003. Exchange Rate Pass-Through To Domestic Prices: The Case Of Colombia. Borradores De Economia 002683, Banco de la Republica Sachs, D.J., 1985. The Dollar and the Policy Mix: 1985. NBER Working Paper Series No. 1636. Sek, S.K. and Kapsalyamova, Z., 2008. Pass-through of exchange rate into domestic prices: The case of four East Asian Countries. The International Journal of Economic Policy Studies, 3 2008 (3). Soto, C. and Selaive, J., 2003. Openness and imperfect pass-through: Implications for the Monetary Policy. Central Bank of Chile Working Papers No.216. Taylor, J.B., 2000. Low Inflation, Pass through and the pricing power of firms. European Economic Review, 44 (2000), 1389-1408.

22