paper

Attribution Non-Commercial (BY-NC)

Als PDF, TXT **herunterladen** oder online auf Scribd lesen

7 Aufrufe

paper

Attribution Non-Commercial (BY-NC)

Als PDF, TXT **herunterladen** oder online auf Scribd lesen

- Sysco Weekly Market Trends Report 11/9/2012
- Solutions Chapters 10 & 11 Transactions and Economic Exposure
- ADFINAP
- midterm.doc
- Assignment # 1, International Finance
- GS Gold BuyRec Comment
- A Study on Financial Derivatives (Futures & Options)
- Sugar Jan 1 2010
- RD
- portfolio
- Euro Money Article 101109
- Indian Commodity Futures Markets
- Inrernationa Monetary Fund
- Forex Mgmt
- Basics
- Top50_2010
- MCom06
- Futures and Options Trading for Hedge Funds. the Regulatory Article)
- Derivatives v 2
- 43747635

Sie sind auf Seite 1von 33

Nongnuch Tantisantiwong School of Business University of Dundee Dundee DD1 4HN UK email:n.tantisantiwong@dundee.ac.uk tel: +44 1382 385838 January 20, 2011

Abstract The framework presents how trading in the foreign commodity futures and domestic forward foreign exchange markets can aect the optimal spot positions of domestic commodity producers and traders. It generalizes the models of Kawai and Zilcha (1986) and Kofman and Viaene (1991) to allow both intermediate and nal commodities to be traded in the international and futures markets, and the exporter to face production shock, domestic factor costs and a random price. Applying the mean-variance expected utility, we nd that a rise in exchange rate volatility can reduce both supply and demand for commodities and increase the domestic prices if the exchange rate elasticity of supply is greater than that of demand. Even though the forward foreign exchange market is unbiased, and there is no correlation between commodity prices and exchange rates, the exchange rate can aect domestic trading and prices through oshore hedging and international trade if the traders are interested in their prot in domestic currency. It illustrates how the world prices and foreign futures prices of commodities and their volatility can be transmitted to the domestic market as well as the dynamic relationship between intermediate and nal goods prices. The equilibrium prices reect trader behaviour i.e. who trade or do not trade in the foreign commodity futures and domestic forward currency markets. The empirical result applying a two-stage-least square approach and Thai rice and rubber prices supports the theoretical result. Keywords: oshore hedging, currency hedging, asset pricing, price transmission JEL Classication: F1; F3; G1; Q1

1

Electronic copy available at: http://ssrn.com/abstract=1744049

Price Transmission and Eects of Exchange Rates on Domestic Commodity Prices via Oshore Hedging

Abstract The framework presents how trading in the foreign commodity futures and domestic forward foreign exchange markets can aect the optimal spot positions of domestic commodity producers and traders. It generalizes the models of Kawai and Zilcha (1986) and Kofman and Viaene (1991) to allow both intermediate and nal commodities to be traded in the international and futures markets, and the exporter to face production shock, domestic factor costs and a random price. Applying the mean-variance expected utility, we nd that a rise in exchange rate volatility can reduce both supply and demand for commodities and increase the domestic prices if the exchange rate elasticity of supply is greater than that of demand. Even though the forward foreign exchange market is unbiased, and there is no correlation between commodity prices and exchange rates, the exchange rate can aect domestic trading and prices through oshore hedging and international trade if the traders are interested in their prot in domestic currency. It illustrates how the world prices and foreign futures prices of commodities and their volatility can be transmitted to the domestic market as well as the dynamic relationship between intermediate and nal goods prices. The equilibrium prices reect trader behaviour i.e. who trade or do not trade in the foreign commodity futures and domestic forward currency markets. The empirical result applying a two-stage-least square approach and Thai rice and rubber prices supports the theoretical result. Keywords: oshore hedging, currency hedging, asset pricing, price transmission JEL Classication: F1; F3; G1; Q1

Introduction

Many frameworks try to explain the behaviour of traders and commodity prices in an economy that has both commodity spot and futures markets. However, some small countries are the world largest exporters or importers of commodities, and they do not have their own commodity futures market. The domestic traders in these small countries are price takers. The world prices can therefore aect the domestic commodity prices, and the higher volatility of world prices can have an adverse eect on these small economies. For example, many oil importing countries suered from the recent oil price crisis in 2008. The ination of South Korea, Taiwan and the Philippines rose sharply due to the rise in oil and rice prices as they are the world largest importers of both commodities. On the other hand, the higher volatility of oil and rice prices causes economic instability of oil and rice exporting countries. For instance, in the rst half of 2008 Thai people suered from a continuous increase in the domestic price of rice which is main and necessary food of the country because more rice was exported due to the higher world price, leaving the country with the lower supply. The rise in oil price caused a higher production costs of synthetic rubber and thus increased the demand for substitutes (e.g. natural rubber). Thailand which is the world largest exporter of milled rice and rubber products beneted from this. That is, Thai export growth increases, so Thailand experienced a resilient economic growth in spite of facing higher fuel import bills. Later, the commodity prices dropped 1

Electronic copy available at: http://ssrn.com/abstract=1744049

following a sharp fall in oil price in July 2008. This caused a decrease in Thai export growth, farm incomes and incomes of rice mills and rubber sheet manufacturers, leading to a reduction of private consumption growth. As a result, the economic growth dropped from 0.8% y-o-y in the rst half of 2008 to 0.7 and -4.9% y-o-y in the third and fourth quarters of 2008, respectively. From this, we can see that traders facing price risk are not only exporters, but also producers, processors and storage companies. Without the domestic futures market, domestic traders can only hedge their price risk in the foreign commodity futures markets or use the foreign futures prices as information in predicting future commodity prices. For example, before the Agricultural Futures Exchange of Thailand (AFET) started trading futures contracts of smoked rubber sheet in May 2004 and futures contracts of milled rice in August 2004, Thai rubber traders hedge their price risk in the Singapore Commodity Exchange (SICOM)1 and Thai rice traders trade in the Chicago Board of Trade (CBOT)2 . This raises a concern whether exchange rates can aect domestic commodity prices also through foreign futures trading. Kawai and Zilcha (1986) found that if the economy has only the forward foreign exchange market, exports could be increased by the introduction of domestic commodity futures markets and decreasing exporters risk aversion. Many recent literatures found the optimal oshore hedging strategy for the exporters or the importers. Kawai and Zilcha (1986) and Kofman and Viaene (1991) found the exporters optimal strategy in the case of incomplete market such that there is no commodity futures market in the economy. Their models were focused on intermediate commodity which was storable, quoted in the foreign demanding countrys currency and traded in the international and futures markets. Yun and Kim (2010) found that Korean oil importers hedging would be more eective if they hedged their price risk in the foreign futures market and simultaneously entered into currency futures contracts. Moreover, Jin and Koo (2002) developed a theoretical model to nd a hedging strategy when the traders face hedging cost and in their empirical work they also found the optimal hedging ratio for Japanese wheat importers who hedged their price risk in the CBOT and hedged exchange rate risk in the Tokyo International Financial Futures Exchange (TIFFE). Unlike others, Kofman and Viaene (1991) also found the optimal strategies of domestic producers and processors as well.

1 There are more Thai traders trading the futures contract of smoked rubber sheet in the SICOM than in other futures market because the Singapore is the largest port shipping smoked rubber sheet for Thailand and Malaysia, the world largest exporters of smoked rubber sheet. 2 For the case of rice, there are more Thai traders trading in the CBOT than in other futures market because the US is one of the largest importers of Thai milled rice. However, rice traded at the CBOT is rough rice, not milled rice.

2

Electronic copy available at: http://ssrn.com/abstract=1744049

While these models focused on optimal strategies of exporters of intermediate commodities, some small countries allow exports of nal goods only e.g. Thai rough rice and natural rubber cannot be exported due to nature of products, much higher delivery cost, and regulations. For some commodities e.g. sugar and oil, both intermediate and nal commodity (raw sugar and rened sugar, or crude oil and rened oil) can be exported in some countries. Some literatures found the empirical relationship between spot exchange rates and domestic commodity prices. Investigating the relation between Thai milled rice price and the US dollar exchange rate, Kofman and Viaene (1991) found a positive ex-post correlation coecient while Gilbert (1991) found the exchange rate elasticity equal to -1. Applying Granger Causality test, Timmer (2009) found that the Euro-US Dollar exchange rate signicantly Granger caused prices of commodities e.g. rice, corn, wheat, crude oil and palm oil. In addition, Chen, Rogo and Rossi (2010) showed that exchange rates had signicant power in forecasting commodity prices. Apart from commodity prices, some studies currency hedging for international nancial portfolio; for example, Schmittmann (2010) found that both exchange rate volatility and the correlations of exchange rates with bond and equity returns can increase risk exposure to investors. An aim of this paper is to develop a framework to explain what are the eects of trading in the foreign commodity futures and domestic currency markets on the domestic commodity market of a small country and how exchange rates can aect domestic commodity prices through this trading. The framework here expands the models of Kawai and Zilcha (1986) and Kofman and Viaene (1991), allowing either intermediate commodity or nal commodity to be traded in the international and futures markets. Some of their assumptions are also relaxed e.g. exporters face export uncertainty, domestic factor costs and domestic price of nal goods. By applying a two-period mean-variance approach, the agents optimal commodity spot and futures holdings and optimal forward exchange holding in this particular case are specied. Unlike Danthine (1978), Holthausen (1979), Feder, Just and Schmitz (1980), and Benninga and Oosterhof (2004), the degree of risk aversion can aect all optimal spot positions. We nd that relationships among optimal strategies, equilibrium prices, and exchange rates depend upon which product is exported, whether exports are contracted in advance, which traders hedge their price and exchange rate risks, and whether the commodity prices and the exchange rates are correlated. We show that even if there is no correlation between commodity prices and

exchange rates, there are still some exchange rate eects on the commodity prices via exporting/importing and trading in the foreign futures if the agents are interested in their prots in domestic currency. The higher volatility of exchange rate can lower the optimal forward position, and thus the optimal futures and spot position. The higher volatilities of spot prices and futures price will reduce both demand and supply of the commodity, but their impacts on price depending on which between supply and demand is aected more. IAs another aim of the paper, the framework shows how world prices are transmitted to domestic intermediate and nal commodity markets. For empirical studies, we use the data on Thai rice and rubber prices, and the exchange rates of Thai Baht against US Dollar and Singapore Dollar before futures contracts rst traded in the AFET. The empirical nding supports the theoretical result. That is, the contemporaneous correlations of exchange rates with commodity prices are insignicant for the case of rice and signicant for the case of rubber. This result is dierent from the ndings of Kofman and Viaene (1991) and Gilbert (1991); it may be due to Thailand moving from the x exchange rate system to the managed-oat system in 1997. The forward foreign exchange market is biased, allowing the forward exchange rate to have impacts on the optimal spot position and commodity prices. As an advantage of this framework, the estimation result can reect who trades in the foreign commodity futures and forward foreign exchange markets and who does not. The rest of paper is outlined as follows. In the next section, the theoretical framework is developed to nd the optimal strategies and equilibrium prices. The robustness of the theoretical result is also discussed. In Section 3, the two-stage-least-square regression shows how prices of intermediate and nal commodities are related and how they are aected by world prices, foreign future prices and exchange rates. Section 4 concludes and suggests some further research.

2

2.1

Framework

Assumptions

In this framework, both intermediate and nal commodities are storable and internationally tradable. Here, we consider how oshore commodity hedging and domestic currency hedging of traders can aect the domestic spot commodity market of a small country which does not have its own commodity futures market, but has

Producer

Consumer

Storage companies

Flows of Primary Commodity Flows of Intermediate Commodity Flows of Final Commodity

Exporters

World Market

Figure 1: Flows of intermediate and nal commodities within an exporting country the forward foreign exchange market. The mean-variance expected utlity maximization is applied to nd the optimal strategies of domestic traders: producers, processors, exporters (or importers), and storage companies. Production and trade ows among traders for the commodities that the country exports is shown in gure (1) while ows for the commodities that the country imports is shown in gure (2) We assume that the production process, storage and internationally delivery take one period (i.e. traders hold the positions for a single-period horizon), and in each period both intermediate and nal commodities are traded in the domestic market. The intermediate producers choosing the optimal level of input (e.g. primary commodity) at period t and sell their output (intermediate commodity) to processor, storage companies or exporters at price Pt+1 at period t+1. With the optimal level of input (intermediate commodity) chosen at period t, processors produce the nal commodity and sell it to storage companies or exporters at price Qt+1 at period t+1. Storage companies can buy either intermediate or nal commodities from the domestic markets and sell them to the domestic market in the next period to make a prot. For the exporting country, exporters buy the commodities from the domestic market; after packing they deliver the commodities to the foreign importers and get paid in foreign currency in the next period at the world prices (Pmt+1 and Qmt+1 ). In the case that the country imports commodities, importers pay for importing commodities at the world prices at period t. They sell the commodities to the domestic market in the next period after receiving

Producer

Consumer

Storage companies

Flows of Primary Commodity Flows of Intermediate Commodity Flows of Final Commodity

Importers

World Market

Figure 2: Flows of intermediate and nal commodities within an importing country commodities. The export/import and futures prices are quoted in the foreign demanding countrys currency. As the country does not have a commodity futures market, domestic traders can hedge their price risk in the foreign commodity futures market, and hedge their exchange rate risk in the domestic forward exchange market. Either intermediate or nal commodity is traded in the futures market. In addition, the maturity of the futures and forward contracts is at time t + 1. Note that in period t all stochastic variables of time t + 1 are unknown. In the sequel, Xit denotes the position of the agent of type i in the primary good at time t and similarly for Yit (intermediate good), Hit

f (nal good) and Yit (futures contract for intermediate or nal commodity). i is the prot of trader i where

i = f for the intermediate producers, i = p for the processors, i = sy for the companies storing intermediate good, i = sh for the companies storing nal good, i = ey for the rm exporting intermediate good and i = eh

f denote the position of the agent of type i in the forward exchange for the rm exporting nal good. Let Zit

contract at time t. The framework also has assumptions as follows. A1. All agents can hedge or speculate in the foreign futures market by selling or buying the futures

f Ft et in domestic currency. It follows that all contract with a full margin at time t which is worth Yit

domestic traders close their futures position by the last trading day of period t+1 by a cash settlement

because delivering to or taking delivery from the foreign futures market requires substantially additional costs, or the commodity traded in the foreign futures market is dierent from the commodity of which price is hedged. All traders transfer the cash (in foreign currency) through the foreign futures markets

f Ft is made at time t with the exchange rate clearinghouse both at time t and t + 1 e.g. the payment on Yit f Ft+1 is made at time t+1 to close futures position with the exchange rate et+1 . et and the payment on Yit

A2. Due to international trade and foreign futures trading, traders face exchange rate risk. To hedge exchange rate risk, traders can short or long the forward exchange contract at time t, based on the expected future payment or receipt (in foreign currency). Unlike the trade in the futures market, there is no margin requirement in the forward market so the payment in the forward exchange market is only made at maturity t + 1. After the cash transfer, the foreign currency remaining in their hands can be sold to the spot exchange

f , they can also buy market. If their actual payment (receipt) in foreign currency is larger (smaller) than Zit

more foreign currency in the spot market. A3. At time t, all traders choose their optimal decisions by maximising their expected utility function (V ) depending on their prot in domestic currency. Any variables such as the optimal spot and futures positions at time t chosen by agent i at time t depend on his own information set available at time t (Iit ). Eit () denotes agent is expectation depending on Iit . V arit () denotes agent is expected variance of a variable depending on Iit . The discount factor, , is assumed to be 1/(1 + it+1 ) where it+1 is the interest rate at time t+1 and perfectly foreseen at time t. A4. Domestic traders in a commodity market are rational and are small in the international commodity market, foreign futures market and domestic forward exchange market relatively to trading volume in the markets. So Eit (et+1 ) = Et (et+1 ). ft = Et (et+1 )+ risk premium. This risk premium can be time varying. Unlike the assumption of Battermann, Braulke, Broll and Schimmelpfennig (2000) here the expected future exchange rate for trader i, Eit (et+1 ), is not necessarily equal to the forward exchange rate, ft . A5. Production shock, storage and export uncertainty, noise trading in the commodity futures and forward foreign exchange markets are uncorrelated and do not have serial correlations. The domestic spot and forward exchange markets are insignicantly aected by the uncertainty of commodity spot and futures prices.

2.2

2.2.1

Prot Functions

Intermediate producers

A producer buys primary commodity at time t to produce intermediate commodity and sells his output (f (Xf t , l,

t+1 )) 3

in the spot market at time t +1. We assume that the production shock,

t+1

has Ef t (

t+1 )

0 and V arf t (

t+1 )

= 2 . The production shock is realised just before delivering the output to the spot market

2 at time t + 1. Xf t is the production cost excluding the cost of seeds ( > 0). He can hedge his price risk

by selling futures contracts maturing at time t + 1, Yfft , at price Ft in the foreign futures market at time t. His prot function is, therefore,

2 f = Yfft t rt Xf t Xf t + {(f (Xf t ) + f Yfft t+1 + (et+1 ft )Zf t }.

t+1 )Pt+1

(1)

where rt is the primary commodity price and t (=Ft et ) is the foreign futures price in domestic currency at time t. Because of closing futures position with cash settlement, he faces exchange rate risk which he can hedge by buying forward exchange contracts maturing at time t + 1 at the forward rate ft . He may buy the exact amount of the foreign currency which he has to pay to the futures market from the spot exchange

f market at t +1 at the rate et+1 and thus Zf t = 0. Alternatively, he may take delivery of the amount of foreign f f f currency Zf t from the forward exchange market. He may sell his excess foreign currency Zf t Yf t Ft+1 in

the spot exchange market at t + 1 if his actual payment (in foreign currency) to the futures market at time

f t + 1 is smaller than Zf t . On the other hand, he may buy more foreign currency from the spot exchange if f f Zf t < Yf t Ft+1 .

2.2.2

Processors

At time t a processor purchases intermediate goods to produce nal goods. He sells his nal goods at time t+1 in the domestic market of nal commodity at price Qt+1 . At time t, he can also sell futures contracts of

f ) to hedge his price risk. He closes all of his futures position the nal commodity maturing at time t+1 (Ypt

at time t+1 by buying futures contracts at price Ft+1 . He can hedge his exchange rate risk by buying the

f f . After taking delivery of Zpt at the forward rate ft , he may buy the remaining forward exchange contract, Zpt

foreign currency in the spot exchange market if it turns out that he has underhedged his exchange rate risk

3Y 1/2 } + t+1 where l is other inputs and 0 < b < 1. Therefore, the cost function is equal to the cost f t+1 = min{bXf t+1 , l of seeds and the other production cost which depends on the amount of input used in the production.

f f f i.e. Ypt Ft+1 > Zpt . Alternatively, he may choose to buy this amount of the foreign currency, Ypt Ft+1 , in f = 0. A processors prot function is the spot exchange market at the rate et+1 at time t+1 and thus Zpt

f f f 2 t Pt Ypt Ypt + {(g (Ypt ) + vt+1 )Qt+1 Ypt p = Ypt +1 t+1 + (et+1 ft )Zpt }

(2)

where g (Ypt , k, t+1 ) is a production function of the nal good (Hpt+1 )4 . The production shock, t+1 has a

2 zero mean and a constant variance ( 2 v ). Ypt is the other production cost which depends on the amount of

A storage company purchases intermediate goods or nal goods at time t and sells them to make a prot

y 2 2 h or h Hsh,t ) where 0 < < 1. wt at time t+1. His storage cost is ( y Ysy,t +1 and wt+1 denotes storj j 2 age uncertainty for intermediate goods and nal goods with Esj,t (wt +1 ) = 0, V arsj,t (wt+1 ) = jw , and y h Covsj,t (wt +1 ,wt+1 ) = 0 where j = y and h. At time t, he can hedge his price risk by selling the futures

contract maturing at time t and close the futures position at time t+1. In addition, he can hedge his exchange rate risk due to the payment to the futures market at time t+1 with the forward exchange contract

f f f ) at the forward rate, ft . He may sell Zsj,t Ysj,t Ft+1 in the spot exchange market if it turns out (Zsj,t

that he has overhedged his exchange rate risk. Instead, he may choose to buy foreign currency from the

f = 0. The prot of the company that stores the spot exchange market at t+1 at the rate et+1 i.e. Zsj,t

intermediate commodity is

f y f f 2 t Pt Ysy,t y Ysy,t + {(Ysy,t + wt sy = Ysyt +1 )Pt+1 Ysy,t t+1 + (et+1 ft ) Zsy,t }.

(3)

For the company that stores the nal commodity, the companys prot function is

f f f 2 h t Pt Hsh,t h Hsh,t + {(Hsh,t + wt sh = Ysh,t +1 )Pt+1 Ysh,t t+1 + (et+1 ft ) Zsh,t }.

(4)

2.2.4

Exporters/Importers

An exporter commits to export intermediate goods or nal goods. He purchases the commodities at time t and export them at time t+1. Like Kawai and Zilcha (1986), we assume that export prices at time t+1 are random. His total production cost is the cost of commodity (Pt Yey,t or Qt Heh,t ) plus delivery and

4H 1/2 } + pt+1 = min{a(Ypt ), k t+1 where k is other inputs and 0 < a < 1. Therefore, the cost function is equal to the cost of intermediate goods and the other production cost which depends on the amount of intermediate goods used in the production.

2 2 h transaction costs (Yey,t or Heh,t ) where 0 < j < 1. uy t+1 and ut+1 denotes uncertainty aecting exports j y 2 h of intermediate goods and nal goods with Eej,t (uj t+1 ) = 0, V arej,t (ut+1 ) = ju and Covej,t (ut+1 ,ut+1 ) = 0

where j = y and h. At time t, he can also hedge his price risk in the foreign futures market by selling the futures contract maturing at time t and close it at maturity. In addition, he can hedge his exchange rate risk due to the payment to the futures market at time t+1 by purchasing the forward foreign exchange

f f f ) at time t at the forward rate ft . He may sell Zej,t Yej,t Ft+1 in the spot exchange market contract (Zej,t

if it turns out that he has overhedged his exchange rate risk. Or else, he may buy foreign currency from the

f = 0. spot exchange market at t+1 at the rate et+1 so Zej,t

f f f 2 t+1 Pt Yey,t Yey,t + {(Yey,t + uy ey = Yey,t t+1 )Pmt+1 et+1 Yey,t t+1 + (et+1 ft ) Zey,t }.

(5)

f f f 2 t Qt Heh,t Heh,t + {(Heh,t + uh eh = Yeh,t t+1 )Qmt+1 et+1 Yeh,t t+1 + (et+1 ft ) Zeh,t }.

(6)

In the framework of an imported commodity, the importers input costs will be the world prices, Pmt and Qmt , instead of Pt and Qt while the commodity is sold at time t+1 at the domestic price. Given that he makes decision at time t, he faces price risk from selling the commodity to the domestic market and also exchange rate risk from oshore hedging.

2.3

Optimal Strategies

Each trader chooses his optimal commodity spot position, commodity futures position and forward exchange position by maximising the mean-variance expected utility depending on their prots. The decision is made at time t, so the domestic supply of intermediate and nal commodities at time t+1 depends on the input that producers, processors and storage companies bought at time t. The objective function

(7)

is maximised with respect to the spot position (Xf t for i = f , Yit for i = p, sy and ey , and Hit for i = sh

f f ) and the forward position (Zit ). and eh), the futures position (Yit

10

2.3.1

The optimal spot position of the producer is Xf t while the optimal spot position for the processor is Ypt . From this, we also nd the optimal output level at time t+1 regardless of production shocks i.e. f (Xf t) ) for the processor. For storage companies, the optimal spot position is trading for the producer and g (Ypt

volume, not the stock of commodity, at time t which is carried forward to the period t+1. The optimal

and for the company storing position at time t for the company storing intermediate goods is denoted as Ysy,t . Yey,t is the optimal position at time t for the exporter of intermediate goods and Heh,t nal goods as Hsh,t for the exporter of nal goods. Let Sit be the vector of the optimal spot positions i.e. f (Xf t ) for i = f , ) for i = p, Yit for i = sy, ey and Hit for i = sh, eh. g (Ypt

Solving the rst order conditions (see Appendix A) yields the optimal position

Sit

Eit (Mit+1 ) Cit Ai Eit (Mit+1 )Covit (Mit+1 , it+1 ) = it s t Eit (t+1 ) V arit (Mit+1 ) Corrit (Mit+1 , t+1 ) Corrit (t+1 , et+1 )Corrit (Mit+1 , et+1 ) + 2 ( it V arit (t+1 ) 1 Corrit t+1 , et+1 ) s (Eit (et+1 ) ft ) V arit (Mit+1 ) Corrit (t+1 , et+1 )Corrit (Mit+1 , t+1 ) Corrit (Mit+1 , et+1 ) + (8) 2 ( it V arit (et+1 ) 1 Corrit t+1 , et+1 )

2 (t+1 , et+1 ) > 1 > Corrit

Mit+1 is the price of product sold by trade type i: Pt+1 for the producer and the company storing or importing the intermediate commodity, Qt+1 for the processor and the company storing or importing the nal commodity, Pmt+1 et+1 for the exporter of the intermediate commodity, and Qmt+1 et+1 for the exporter of the nal commodity. Cit is the input cost: rt /b for the producer, Pt /a for the processor, Pt for the company storing or exporting intermediate goods, Qt for the company storing or exporting nal goods, Pmt+1 et+1 and Qmt+1 et+1 for the importers of intermediate and nal commodities. it+1 is production shock for the producer and the processor, storage uncertainty for storage companies, and export (import) uncertainty for exporters (importers). We assume that Covit (Mit+1 , it+1 ) < 0. This is because a decrease in the supply of

11

the commodities due to a negative shock (e.g. weather and rotting) will raises the commodity price. it = i + Ai V arit (Mit+1 ){1

2 2 Corrit (Mit+1 , t+1 ) + Corrit (Mit+1 , et+1 ) 2 ( 1 Corrit , e t+1 t+1 ) 2Corrit (Mit+1 , et+1 )Corrit (t+1 , et+1 )Corrit (Mit+1 , t+1 ) + } 2 ( 1 Corrit t+1 , et+1 ) 2 a2

where i is equal to

2 b2

for the exporters (importers). i = 0 if there is no other costs apart from the cost of commodity. it > 0 implies that the second order conditions (SOCs) are satised. Unlike Anderson and Danthine (1983), Antoniou (1986), and Benninga and Oosterhof (2004) in which the country has its own futures market, the optimal spot position here depends on trading in other markets and the degree of risk aversion; thus, the separation theorem is not applicable. When the expected gain in the spot, futures or forward market increases, the traders tend to hold a larger spot position. Like Kawai and Zilcha (1986), Kofman and Viaene (1986) and Schmittmann (2010), the failure of uncovered interest rate parity (UIP) allows currency hedging to have an impact on the spot position. While the higher volatility of domestic spot prices, world prices and foreign futures prices of commodities and exchange rate can cause a reduction of the optimal spot positionm, the higher Corrit (Mit+1 , t+1 ) reduces the optimal spot position (See Appendix B). This is supported by the nding of Bahmani-Oskooee and Mitra (2008) that exchange rate volatility has negative impacts on exports and imports of commodities between the US and India. The eects on his prot are in the same direction as those on the optimal spot positions if the expected marginal gain of trading in futures and forward markets is non-negative and that the marginal revenue product (MRP) of input is not less than its marginal resource cost (MRC). If there is no correlation between commodity prices and exchange rates, Corrit (Rt+1 , et+1 ) = 0 for all i where R = P, Q, Qm , Pm , F. For Lt+1 = t+1 , Pmt+1 et+1 and Qmt+1 et+1 , Corrit (Lt+1 , et+1 ) = = = Covit (Lt+1 , et+1 ) p V arit (Lt+1 )V arit (et+1 ) Eit (Lt+1 /et+1 )V arit (, et+1 ) + Eit (et+1 )Covit (Lt+1 /et+1 , et+1 ) p V arit (Lt+1 )V arit (et+1 ) s V arit (et+1 ) Lt+1 Eit (Lt+1 /et+1 )V arit (, et+1 ) p = Eit ( ) et+1 V arit (Lt+1 ) V arit (Lt+1 )V arit (et+1 )

Eit (t+1 ) = Eit (Ft+1 )Eit (et+1 ) for all i and Eit (Mit+1 ) for the exporters are Eey,t (Pmt+1 )Eey,t (et+1 ) and Eeh,t (Qmt+1 )Eeh,t (et+1 ). Then, the optimal positions of traders are still aected by the exchange rate and 12

its volatility. If the forward market is also unbiased, or Corrit (Mit+1 , et+1 ) = 0 and Corrit (t+1 , et+1 ) = 0 for all i, the optimal position will be Sit Eit (Mit+1 ) Cit Ai Eit (Mit+1 )Covit (Mit+1 , it+1 ) = 2 (M i + Ai V arit (Mit+1 ) [1 Corrit it+1 , t+1 )] s V ar (M [t Eit (t+1 )] Corrit (Mit+1 , t+1 ) it it+1 ) + 2 (M i + Ai V arit (Mit+1 ) [1 Corrit V arit (t+1 ) it+1 , t+1 )]

(9)

This optimal spot positions is not aected directly by the exchange rates but indirectly through foreign futures trading and exports (see Appendix C). The second term of the optimal input level of (8) and (9) is the eect of futures trading. The sum of the rst two terms of (8) is larger than (9). With the eect of speculative trading in the forward foreign exchange market on the optimal spot position, (8) (9). (9) is also the optimal position for the case that 1) the economy does not have a forward foreign exchange market or 2) the trader hedges only his price risk. If the producer does not hold futures position or if he is not allowed to trade in the foreign futures market, he does not need to hedge exchange rate risk. In this case, the optimal spot position (8) will become Eit (Mit+1 ) Cit Ai Eit (Mit+1 )Covit (Mit+1 , t+1 ) Sit = (Ai V arit (Mit+1 ) + i ) (10)

for all i. If the trader does not trade in the foreign futures market but use the futures price to predict the future spot price, Eit (Pt+1 ), Eit (Qt+1 ), Eey,t (Pmt+1 ) and Eeh,t (Qmt+1 ) in (10) will be replaced by or be a function of Ft . As the optimal spot position (10) <(9) <(8), allowing the trader to hedge his risks in the foreign futures market and the domestic forward exchange market will raise production level and supply of commodity in the domestic market. By how much the production level or trading volume in the spot market for each trader increases depends on whether the commodity prices and foreign exchange rates are correlated and whether the forward exchange market is unbiased. If the nal good production process is short, the processor can buy the intermediate commodity to produce the nal good and sell the output to the market at time t. In this case he does not face any price risk and has no need to trade in the foreign futures and forward foreign exchange markets. His optimal spot position is Ypt = where gyp =

g (Ypt ) Ypt .

(11)

Unlike other trader types, exporters face exchange rate risk even though they do not 13

involve in the foreign commodity futures market. If they choose to hedge their price risk in the oshore futures market, then they face more exchange rate risk. As can be seen from equation (8), the optimal spot position depends on whether he also trades in the foreign futures market and the forward exchange market. If he does not trade in the futures market, but he hedges the exchange rate risk of his export incomes, Corrit (t+1 , et+1 ) = 0 and Corrit (Mit+1 , t+1 ) = 0. Thus, the second term disappear and the last term is reduced; therefore, his optimal spot position becomes smaller than (8) but greater than (10). Considering a case in which the exporter precommits to export intermediate (nal) commodity at time t at the preset export price Pmt+1 (Qmt+1 ), the preset export price can be determined by the current export price and thus Pmt+1 (Qmt+1 ) are known at time t as assumed by Kofman and Viaene (1991) and Benninga and Oosterhof (2004). The exporter has no need to hedge price risk. The optimal positions of the exporters will be Yey,t = and Heh,t = 2.3.2 The optimal futures position [Qmt+1 ft Qt ] 2 (13) [Pmt+1 ft Pt ] 2 (12)

f Yit

for all i. The optimal futures position has 2 components: speculation and hedging. The rst two terms are eects of speculation in the foreign commodity futures market and the forward foreign exchange market.

H is the hedging component: Yit s V ar (M ) {Corr (M it t+1 it it+1 , t+1 ) Corrit (Mit+1 , et+1 )Corrit (t+1 , et+1 )} H Yit = Sit 2 ( V arit (t+1 ) (1 Corrit t+1 , et+1 ))

(t Eit (t+1 )) 2 ( Ai V arit (t+1 ) ((1 Corrit t+1 , et+1 ))) (Ef t (et+1 ) ft )Corrf t (t+1 , et+1 ) H + + Yit p 2 ( Ai V arit (t+1 )V arit (et+1 ) (1 Corrit , e )) t+1 t+1

(14)

H for all i. Yit is a partial hedge of the expected output level. Whether the trader will short or long the

contract depends on the expected gains in the foreign futures and forward foreign exchange markets, and his spot position. As

2 (t+1 , et+1 ) > Corrit

if there is no correlation between commodity prices and exchange rates and the trader hedges his exchange rate risk, the denominator increases as

2 V arf t (t+1 Eit (Ft+1 )et+1 ) > V arf t (t+1 )(1 Corrf t (t+1 , et+1 )) > 0.

and Covit (t+1 , et+1 ) = Eit (Ft+1 )V arit (et+1 ). Thus, the optimal futures position will be smaller than (14).

f is zero for all i. Then the second term disappear and If traders do not hedge their exchange rate risk, Zit

there are no Covit (t+1 , et+1 ), Covit (Pmt+1 et+1 , et+1 ), Covit (Qmt+1 et+1 , et+1 ) in (14). The optimal position is reduced to be

f = Yit

(15)

H in (15) is greater than in (14) i.e. oshore hedging is more eective when they hedge both We nd that Yit

price and exchange rate risks. This is supported by empirical ndings of Yun and Kim (2010). Suppose that only intermediate commodity is traded in the foreign futures market. The size of futures position may be aected by the type of commodity traded in the foreign futures market through the futures price and Covit (Mit+1 , t+1 ). For example, the covariance between the domestic price of nal (intermediate) good and the foreign futures price of intermediate (nal) good may be lower than the covariance between the domestic prices and the foreign futures price of the same good. The optimal spot and forward positions are also aected through the change in futures position. 2.3.3 The optimal forward position

f = Zit

Eit (et+1 ) ft f Covit (t+1 , et+1 ) Covit (Mit+1 , et+1 ) + Yit Sit Ai V arit (et+1 ) V arit (et+1 ) V arit (et+1 )

(16)

f is composed of two components: speculation (the rst term) and hedging (the last two Obviously, Zit

terms). The second term of (16) is a partial hedge of the expected payment of receipts of foreign currency due to closing futures position at time t+1 and its last term is a partial hedge of the expected future receipts from the spot market. If commodity prices and exchange rate are uncorrelated, the second term

f Eit (Ft+1 ), and the last term will be zero for all traders except exporters. This will become a full hedge, Yit

is because the output prices for exporters are Pmt+1 et+1 and Qmt+1 et+1 i.e. Covey,t (Pmt+1 et+1 , et+1 ) and

15

Coveh,t (Qmt+1 et+1 , et+1 ) in this case are equal to Eey (Pmt+1 )V arey,t (et+1 ) and Eey (Qmt+1 )V arey,t (et+1 ), respectively. In short, the optimal spot position is aected by the degree of risk aversion and thus the separation theorem is not applicable. Without hedging price risk in the foreign futures markets or exchange rate risk in the forward foreign exchange markets, the spot commodity markets will have both lower demand and lower supply. For all traders, all optimal positions are aected by the type of commodity traded in the foreign futures market only through the futures prices and covariance between output prices and foreign futures prices. Note that by changing the subscript t to be t+1, we can get the optimal positions at time t+1 for all traders.

2.4

Equilibrium solution

From the optimal spot positions in the previous section, the market-clearing conditions for the markets of intermediate and nal commodities in the exporting country at time t+1 are

t+1

y + wt + nsy [Ysy,t +1 ]

(17)

where Hct +1 (Qt+1 ) is the demand for nal goods of consumers at time t+1, assuming to be linear with

h neh Heh,t +1 + nsh Hsh,t+1 + Hct+1 = np aYpt + t+1 + nsh Hsh,t + wt+1

(18)

the current domestic price. ni denotes the number of the trader type i. The left-hand side represents the demand in the spot market while the right-hand side represents the supply5 . As domestic traders are small in the international market and the foreign futures and forward foreign exchange markets, the futures price, export prices and spot and forward exchange rates are exogenous. With both equilibrium conditions, the domestic equilibrium spot prices of intermediate and nal commodities at time t+1 are specied. The supply of commodites does not depend on the current domestic spot prices. It depends on the domestic spot prices of inputs, the foreign futures price, and exchange rates at time t, as well as price expectations of producers and storage companies perceived at time t. In contrast, the demand for commodities depends on the current spot price. It also depends on the current futures price

5 For

the importing country, Yey and Heh are on the supply side.

16

and exchange rates, the expected exchange rate, and the expected future prices of commodities sold at time t+2 in the domestic market, the world market and the foreign futures commodity market. So from (17), Pt+1 = nf f (Xf t) + y + np Ypt +1,p + ney Yey,t+1,p + nsy (Ysy,t+1,p Ysy,t + wt+1 ) (19)

t+1

where Yit +1,p denotes the optimal position of the trader type i excluding the eect of Pt+1 . This supports

the assumption given before that Covit (Pt+1 , t+1 ) < 0. Moreover, storage companies or the buer stocks manager can raise the current spot price by buying more of the commodity to increase their stock at time t+1. From (18), the equilibrium price of nal goods in the domestic market is h + t+1 + ne Het Qt+1 = np aYpt +1,q + Hct+1,q + ns (Hst+1,q Hst + wt+1 ) (20)

where Hit +1,q denotes the optimal position of trader type i excluding the eect of Qt+1 . If Hct+1 = cdQt+1 , then Hct +1,q = c.

We can also rewrite (19) as Pt+1 = f (rt , Pt , t , t+1 , ft , ft+1 , it , jt+1 , it , jt+1 ) (nf

y + nsy wt +1 )

t+1

(21)

where i = f, sy and j = p, ey, sy . it and jt+1 are sets of the expected future domestic and world prices of the intermediate commodity (Eit (Pt+1 ) and Eit (Pmt+1 )), and the expected futures price and exchange rate (Eit (t+1 ) and Eit (et+1 )), perceived by the trader type i at time t and the trader type j at time t+1. it and jt+1 are sets of expected variances and correlations, perceived by the trader type i at time t and the trader type j at time t+1, of future prices and exchange rates. (20) can be rewritten as

h Qt+1 = f (Pt , Qt , t , t+1 , ft , ft+1 , kt , mt+1 , kt , mt+1 ) (np vt+1 + nsh wt +1 )

(22)

where k = p, sh and m = e, sh. kt = {Ekt (Qt+1 ), Ekt (t+1 ), Eit (et+1 )} and mt = {Est+1 (Qt+2 ), Eet+1 (Qmt+2 et+2 ), Emt+1 (t+2 ), Emt (et+2 )}. kt and mt+1 are sets of expected variances and covariances of domestic and world prices of nal goods, the futures price and exchange rate, perceived by the trader type k at time t and the trader type m at time t+1. If there is no correlation between prices and exchange rates, with the assumption A4 that Eit (et+1 ) = Ejt (et+1 ) = Et (et+1 ) Pt+1 = f (rt , Pt , t , t+1 , it , jt+1 , it , jt+1 ) (nf 17

y + nsy wt +1 )

t+1

(23)

where it = {Eit (Pt+1 ), Eit (Ft+1 )ft } and jt+1 = {Est+1 (Pt+2 ), Ept+1 (Qt+2 ), Eet+1 (Pmt+2 )ft+1 , Ejt+1 (Ft+2 )ft+1 }.

h Qt+1 = f (Pt , Qt , t , t+1 , kt , mt+1 , kt , mt+1 ) (np vt+1 + nsh wt +1 )

(24)

where kt = {Ekt (Qt+1 ), Ekt (Ft+1 )ft } and mt = {Est+1 (Qt+2 ), Eet+1 (Qmt+2 )ft+1 , Emt+1 (Ft+2 )ft+1 }. As can be seen in (8) and (9), the optimal spot position decreases when V arit (t+1 ) or V arit (et+1 ) increases. Thus the equilibrium price are also aected by exchange rate volatility. However, whether the eect is positive or negative depends on which traders hedge their price and exchange rate risks and the exchange rate elasticity of supply and demand. If the elasticity of supply is greater than, less than, or equal to the elasticity of demand, then prices will increase, decrease or remain unchanged. This is supported by the conclusion of Chu and Morrison (1984) that one of the dominant sources of commodity price variability was the volatility of exchange rates. An increase in the expected value of world prices and a decrease in their volatilities will raise export volume. The increase in export volume causes the domestic demand for commodities, which are the input, to increase; therefore, the domestic prices of commodities increase. In the case that the country is the importer of commodities, if the current world price increases, import volume will decrease. This lowers the domestic supply of commodities and thus increases the domestic price. So the world price and domestic prices are positively correlated.

Empirical studies

Based on the theoretical result in the previous section, trading in the foreign commodity futures and domestic forward exchange markets aects the domestic spot price through the speculative component of the traders optimal spot holding. In this section, the determinants of equilibrium spot prices at maturity are investigated. The aim is to nd whether the spot prices are aected by the foreign futures price and exchange rates through foreign futures trading. This paper applies Thai rice and rubber markets as case studies. As Thailand only

exports nal goods of rice and rubber (milled rice and smoked rubber sheet (RSS)), Yey,t +1 = 0 and the

domestic prices are not aected by the world price of the intermediate good. The export prices of milled rice and RSS3 in this empirical study are Thai F.O.B. prices. For the case of Thai rubber, the AFET has the futures contract of smoked rubber sheet, maturing in September 2004, as the rst product. Due to limitation of data available from the Thai oce of the rubber 18

Table 1: The KPSS unit root Commodity Pt Qt Qmt Qmt et Ft Rubber 0.139 0.131 0.155 0.144 0.151 Rice 0.181 0.175 0.191 0.184 0.186 Critical Values: 1% = 0.216; 5% = 0.146

replanting aid fund and exclusion of the eect of exchange rate regime switching on July 2,1997, the analysis covers the period from January 2001 - June 2004. The domestic prices of intermediate and nal goods applied in this study are of natural rubber sheet and smoked rubber sheet no.3 (RSS3). Rubber tree takes many years to grow, so production cost is mainly xed cost and cost of seed is relatively small and ignorable (rt 0). The futures price of RSS3 is the price of futures contract traded at SICOM. The contract has 12 maturity months i.e. 12 calendar months. Spot and 1-month forward exchange rates are in Thai Baht/Singapore Dollar. So there are totally 42 monthly observations for the case of rubber. Regarding the case of Thai rice, the futures contract maturing in November 2004 is the rst contract of milled rice traded in the AFET. The data covers the period from July 1997 to June 2004. The domestic prices are Thai rough rice and milled rice prices. The primary commodity of rough rice is rough rice, so rt = Pt . The foreign futures price is U.S. rough rice futures price. The CBOT trades rough rice futures contract with 6 maturity months: January, March, May, July, September, and November. The spot and 2-month forward exchange rates are in Thai baht/U.S. dollar. There are totally 46 observations for this case. To simplify the solution and to allow the model to be easily estimated, we can assume that, for each agent, conditional variances and covariances of prices and production shock are constant through time6 . Before estimating the system of equations, the KPSS unit root test is applied to test the stationarity of the variables. The null hypothesis of KPSS test is that the series is stationary. As shown in table 1, all variables are stationary at 0.01 signicance level. So this assumption is valid. Table 2 shows that rubber prices are negatively correlated with the THB/SGD exchange rate while rice prices are not signicantly correlated with the THB/USD exchange rate. Thus, the optimal positions of traders and equilibrium prices in the Thai rubber market and the Thai rice market are dierent. Apart

6 This assumption can be justi ed by a rational expectations paradigm. That is, the conditional distribution for each agent comes from distributions of production shocks and storage uncertainty as well as the market-clearing condition determining Pt+1 and Qt+1 above. Agents know that the domestic commodity spot price is a linear function of production and storage uncertainty. These variables have constant variance-covariance matrices through time so do the distributions it . As Qmt+1 , Pmt+1 , Ft+1 and et+1 are exogenous, the covariance and variances of these variables are given and assumed to be constant. If the mean values of market prices are nonnegative, the rational agents expected values of future prices will also be nonnegative..

19

Table 2: The chi-square test of correlation coecients Pt 1.00 0.999 0.995 0.988 Qt 1.00

(59.958) (41.219) (4.797)

Rubber Qmt et

Ft et

et Pt Qt

Pt Qt Qmt et Ft et et

Qt 1.00 0.985

35.965 (7.009) (0.048)

Rice Qmt et

Ft et

et

0.994 0.988

1.00

(34.904) (5.137)

Qmt et 1.00

(5.017)

1.00

(8.028) (0.064)

0.984

Ft et 1.00 et

0.742 0.008

0.786 0.010

1.00

(0.173)

0.611

0.604

0.630

0.621

0.151

0.027

1.00

Table 3: Market eciency hypothesis testing of forward foreign exchange markets Currency Market THB/SGD Variable Coecient (S.E.) Constant 0.38 (0.19)* ln(ft1 ) 0.88 (0.06)** Adj-R2 84.89% AR Test: F-stat (Prob.) 2.18 (0.13) H0 : 0 = 0, 1 = 1; 2 -stat (prob) 6.52 (0.038) Note: ** and * denote the signicance of coecients at THB/USD Coecient (S.E.) 1.87 (0.42)** 0.50 (0.11)** 30.24% 0.54 (0.59) 9.77 (0.000) 1% and 5% signicance levels, respectively

from the correlation between commodity prices and exchange rate, the commodity equilibrium prices also depend on the biasedness of the forward foreign exchange market. As shown in the theoretical result above and the proposition of Kawai and Zilcha (1986) and Kofman and Viaene (1991), to be persuaded to hold forward contracts, risk-averse traders have to receive a risk premium to compensate for the uncertainty regarding the expected future spot rate. In other words, the forward foreign exchange market is biased i.e. Et1 (ln(et )) 6= ln(ft1 ). The heteroscedasticity-robusted standard error ordinary least square regression of ln(et ) = 0 +1 ln(ft1 )+ t in Table 3 shows that the forward exchange markets are biased and the current exchange rate depends on the previous forward rate. So it is possible to assume that the expected future exchange rate is a function of the current foward exchange rate. Also, trading in the forward exchange market can aect the optimal position in the rubber market and the optimal position is (8) if traders choose to hedge their exchange rate risk. For the rice market, the optimal position is (9). The forward exchange rate is expected to have a signicant eect on Thai rubber prices. Which between intermediate and nal commodity prices is aected by the forward exchange rates depends on who hedge their price risk in the foreign commodity futures market and their exchange rate risk in the forward foreign exchange market.

20

Consequently, we estimate equations (21) and (22) for the Thai rubber market and equations (23) and (24) for the Thai rice market by assuming that all the terms in a coecient (such as the degree of risk aversion, the number of traders and variances and covariances of prices) are constant, Et1 (Pt ) = 0 +

p2 p3 p4 X X X i Qti , Et1 (Qmt ) = 0 + i Qmti , and Et1 (Ft ) = 0 + i Fti . Et1 (Qt ) = 0 + i=1 i=1 i=1 p1 X i Pti and i=1

numbers of lags, pi , are equal to 1. Though more lags may be included if the regressions have autocorrelation problem. Equations (21) and (22) can be rewritten as

p5 p6 X X i Qmti eti and Et1 (Ft et ) = 0 + i Fti eti . Due to the limitation of data, we assume that the 0 + i=1 i=1

For the case of Thai rubber, commodity prices and exchange rates are correlated, so Et1 (Qmt et ) =

Pt = a0 + a1 Qt + a2 Pt1 + a3 Ft1 et1 + a4 Ft et + a5 ft1 + a6 ft + pt . Qt = b0 + b1 Qmt et + b2 Pt1 + b3 Ft1 et1 + b4 Ft et + b5 ft1 + b6 ft + b7 Qt1 + qt . pt = nf

t+1

(25)

(26)

y h + nsy wt +1 and qt = np vt+1 + nsh wt+1 are the residuals. Thus, the mean values of pt and qt

degree of freedom, we assume (ft1 ) = ft1 and pi are equal to 1 and thus equations (23) and (24) can be rewritten as

(ft1 ), For the case of Thai rice prices, Et1 (Qmt et ) = Et1 (Qmt )Et1 (et ) = i =1 ! p4 X i Fti (ft1 ). To simplify the model and avoid losing the Et1 (Ft et ) = Et1 (Ft )Et1 (et ) = 0 +

i=1

p3 X 0 + i Qmti

(27)

Qt = b0 + b1 Qmt ft + b2 Pt1 + b3 Ft1 et1 + b4 Ft et + b5 ft1 + b6 ft + b7 Qt1 + b8 Ft1 ft1 + b9 Ft ft + qt . (28) a0 and b0 are the sums of the constant term in expectation functions, so they can be either positive or negative. Based on the theoretical result, a1 and b1 are expected be positive as Qt and Qmt et are used to forecast the future prices of outputs that have input prices Pt and Qt , respectively. Pt1 is the input cost for farmers and storage companies selling the intermediate commodity at time t and Qt1 is the input cost of storage companies selling the nal good at time t; both are also used to expect the future prices at time t. Therefore, a2 and b7 can be either positive, negative or insignicant. Pt1 is the input cost of processors, so b2 is expected to be positive. While a3 and b3 are expected to be negative as shown in (19) and (20) if

21

Table 4: Regression result of the Thai rubber market Equation Pt Constant Qt Pt1 Ft1 et1 Ft et ft1 ft Ft1 Ft2 et2 Pt11 Pt12 Model RS1 3.17 (7.50) 0.75 (0.23)** 0.05 (0.11) 0.09 (0.18) 0.15 (0.23) -0.02 (0.66) -0.10 (0.62) Model RS2 -0.61 (0.62) 0.96 (0.09)** 0.36 (0.11)** -0.26 (0.15) Model System RS3 -0.68 (0.47) 0.84 (0.05)** 0.18 (0.07)*

-4.69 (1.68)** -0.28 (0.07)** -0.14 (0.08) 0.27 (0.08)** 99.50% -22.61 0.12 (0.88) -22.46 (5.93)** 0.72 (0.19)**

Adj-R2 99.11% 99.38% Log likelihood -50.50 -43.13 Autocorrelation: F stat (prob.) 16.81 (0.00) 4.87 (0.01) Qt Constant -18.34 (8.64)* -12.54 (6.32) Qmt et 0.73 (0.15)** 0.72 (0.12)** Pt1 -0.03 (0.48) Ft1 et1 -0.33 (0.19) -0.35 (0.15)* -0.45 (0.16)** Ft et 0.43 (0.12)** 0.49 (0.10)** 0.52 (0.17)** ft1 1.03 (0.54) 0.48 (0.24) 0.84 (0.23)** ft -0.34 (0.53) Qt1 0.23 (0.46) 0.61 (0.13)** 0.44 (0.13)** Qmt1 et1 -0.48 (0.24) Ft2 et2 -0.32 (0.09)** Adj-R2 99.31% 99.52% 99.48% Log likelihood -46.26 -39.18 -39.80 Autocorrelation: F stat (prob.) 8.02 (0.00) 3.71 (0.04) 2.39 (0.11) Mis-specication: LR stat (prob.) 0.01 (0.91) 0.65 (0.42) 1.00 (0.32) Note: **, *, denote the signicance of coecients at 1%, 5%, 10% signicance levels, respectively producers or storage companies hedge their price risks, a4 and b4 are expected to be positive if processors or exporters hedge their price risk. If the companies storing intermediate goods (nal goods) trade in the futures market, a3 and a4 (b3 and b4 ) both will be signicant. a7 and a8 (b8 and b9 ) should have the opposite sign to a3 and a4 (b3 and b4 ), respectively. If the traders also hedge their exchange rate risk, then a5 , a6 , b5 and b6 of (25) and (26) will be signicant. Applying the two-stage-least-square estimation approach, the estimation result in Table 4 and 5 shows that commodity prices are aected by spot and forward exchange rates through exports and oshore hedging. For the Thai rubber market, as shown in the correlation matrix, the domestic spot price of intermediate goods are highly correlated with the domestic and world prices of nal goods. Thus, the estimators of b2 and b7 in model RS1 are insignicant. Removing Pt1 to correct multicollinearity problem and adding more lags of variables to correct autocorrelation problem yield models RS2 and RS3. The most appropriate

22

Table 5: Regression result of the Thai rice market Equation Pt Model R1 Model R2 Model R3 1126.48 (1183.22) 2323.54 (450.30)** 2107.48 (477.95)** 0.21 (0.07)** 0.24 (0.04)** 0.24 (0.04)** 0.43 (0.18)* 0.31 (0.11)** 0.31 (0.09)** -1.30 (0.72) -1.14 (0.59) -0.98 (0.59) 1.09 (0.69) -33.05 (18.19) -18.83 (9.18)* -14.60 (7.82) 32.13 (27.74) 1.32 (0.71) 1.15 (0.59) 0.99 (0.58) -1.08 (0.66) Adj-R2 85.85% 86.17% 86.30% Log likelihood -289.87 -291.52 -310.78 Autocorrelation: F stat (prob.) 0.34 (0.72) 0.11 (0.90) 0.21 (0.81) Qt Constant -1142.762 (944.62) -1770.13 (260.74)* -1737.10 (1041.59) Qmt ft 0.88 (0.07)** 0.86 (0.05)** 0.91 (0.05)** Pt1 -0.07 (0.10) 0.19 (0.10) Ft1 et1 -0.81 (0.92) Ft et 3.91 (0.54)** 3.81 (0.49)* 3.87 (0.62)** ft1 14.96 (18.62) ft 9.92 (23.37) 31.15 (16.96)* 24.51 (18.32) Ft1 ft1 0.79 (0.90) Ft ft -3.93 (0.53)** -3.85 (0.49)** -3.93 (0.62)* Qt1 0.11 (0.07) 0.13 (0.04)** Adj-R2 97.94% 98.05% 97.65% Log likelihood -284.35 -285.65 -289.55 Autocorrelation: F stat (prob.) 0.61 (0.55) 0.47 (0.63) 0.29 (0.75) Mis-specication: LR stat (prob.) 0.45 (0.50) 2.37 (0.12) 2.41 (0.12) Note: **, *, denote the signicance of coecients at 1%, 5%, 10% signicance levels, respectively model here is RS3 that does not have multicollinearity, autocorrelation and mis-specication problems. The estimators of a1 and b1 are positive as expected. That is, both processors and exporters use the current output price to predict their future output prices. The estimator of a6 is insignicant while the estimator of b5 is signicant; this implies that processors do not hedge their price and exchange rate risks. Thus, the previous forward exchange rate can aect the nal commodity price only through the optimal spot position of storage companies in the previous period. The positive coecient of b5 supports our assumption that

2 1 > Corrit (t+1 , et+1 ) >

and i > 0. The result also shows that exports and storage companies in the nal goods market hedge their price and exchange rate risks; as expected the estimator of b3 is signicantly negative and b4 is signicantly positive. The eects of the current forward exchange rate on the nal goods price through trading by storage companies may be oset by the eects through trading by exporters, causing the insignicance of the estimator of b6 . Unlike the smoked rubber sheet market, traders in the natural rubber sheet market do 23

Actual 60 50 40 30 20 10 0

Jul-03 May-02 May-03 May-04 Mar-03 Mar-02 Nov-02 Sep-02 Nov-03 Sep-03 Mar-04 Jan-02 Jan-03 Jan-04 Jul-02

Actual 60 50 40 30 20 10 0

Mar-02 Dec-01 Dec-02 Mar-03 Jun-02 Jun-03 Jun-01 Mar-01 Dec-03 Sep-03 Sep-01 Sep-02 Mar-04 Jun-04

Fitted

Fitted

Rough Rice Prices

Actual 8000 7000 6000

9000

15000 12000 Actual Fitted

Fitted

5000 4000 3000 Mar-98 Mar-99 Mar-00 Mar-01 Mar-02 Mar-03 Sep-99 Sep-00 Sep-03 Sep-97 Sep-98 Sep-01 Sep-02 Mar-04

6000 3000 Mar-98 Mar-00 Mar-02 Mar-03 Mar-99 Mar-01 Sep-97 Sep-98 Sep-99 Sep-00 Sep-01 Sep-02 Sep-03 Mar-04

Figure 3: Fitted and actual values of intermediate and nal commodity prices not hedge their risks. The negative coecients of futures price Ft1 and Pt11 and the positive coecient of Pt12 indicate that producers use the SICOM futures price to predict their future output price while storage companies use the commodity price at the same period in the year before7 .

For the Thai rice market, removing the insignicant coecients we obtain models R2 and R3: one with the rst lag of intermediate goods price and another with the rst lag of nal goods price in (28). R2 is chosen because all coecients are signicant, the model does not have heteroscedasticity, autocorrelation and mis-specication problems, and it has the higher log likelihood value. The result shows that the producers of intermediate commodity and the exporters of nal commodity hedge their price risk in the CBOT. As expected, the estimators of a1 and b1 are positive. In addition, a3 and b9 are negative while a7 and b4 are positive. These indicate that processors and exporters of nal commodity use the current prices of nal commodity to predict the future price in the domestic and international market. In addition, the signicant estimators of a5 , a7 , b6 and b9 implies that exchange rates have eects on domestic commodity prices through oshore hedging.

7 This re ects the seasonal e ect on natural rubber sheet price. The production level is lower during rainy season. The weather causes disease to rubber tree and reduces the production of rubber latex.

24

Table 6: Descriptive statistics of estimated residuals Statistics Mean Standard Deviation Skewness Kurtosis Jarque-Bera Prob. Rubber Pt Qt 15 2.41 10 1.61 1015 0.523 0.674 0.785 0.471 3.138 3.303 3.108 1.634 0.211 0.442 Rice Pt 1.07 1012 260.063 0.061 2.266 1.038 0.595 Qt 2.66 1013 220.17 0.313 2.960 0.603 0.707

The models have remarkably high explanatory power. The graphs of tted and actual values in gure 3 qt are the estimated residual of which statistics are summarised show that the models t very well. pt and in Table 6. The statistics shows that the estimated residuals are normally distributed with mean values equal to 0 as assumed in the theoretical framework. In short, this empirical result shows that biasedness of the domestic forward exchange market allows the spot and forward exchange rates to aect the commodity prices. This eect depends on whether the commodity prices and exchange rates are correlated and who hedge their price risk in the foreign futures markets and hedge their exchange rate risk in the forward foreign exchange market. Rice producers hedge their price risk while rubber producers do not. This can relate to the commodities that the foreign futures market trades. The CBOT trades futures contracts of rough rice (intermediate goods) while the SICOM trades futures contracts of RSS no. 3 (nal goods). The reason why processors do not hedge their risk may be that the production process take a short period and thus they almost face no price risk; this is consistent with the insignicant coecient of Pt1 in the (26) and (28). Exporters hedge their price and exchange rate risks in both markets. Only the companies storing RSS hedge their price risk; this may be because rough rice is stored by the government agency through price guarantee program while there is no such a program for rubber.

Conclusion

In conclusion, this paper is developed to nd the optimal oshore trading strategy and how it aects the optimal spot commodity and forward foreign exchange positions. It derives equilibrium prices at maturity in the domestic commodity markets to show how oshore trading allows the foreign exchange rates aect the domestic commodity prices in a country without its own commodity futures markets. The framework

25

expands the models of Kawai and Zilcha (1986) to explain trading of other trader types. It relaxes the assumptions of Kawai and Zilcha (1986) and Kofman and Viaene (1991) to allow both intermediate and nal goods to be traded in the international and futures markets. As suggested by Kawai and Zilcha (1986), this framework allows exporters to face export shocks, factor costs and random export prices. Like other frameworks, traders are assumed to close all of their futures position due to high delivery and transaction costs. Applying a two-period mean-variance approach, this framework shows that the changes in the futures price and exchange rates and their volatility can aect the spot positions and domestic prices of internationally tradable commodities in many cases. Firstly, the traders hedge their price risk in the foreign commodity futures market and exchange rate risk in the forward foreign exchange markets which are not unbiased and the commodity prices are correlated with the exchange rate. Second, exporters, importers and the traders hedging their price risk in the foreign futures market are interested in the prots in domestic currency; the eects exist even though the commodity prices and exchange rates are uncorrelated. Thirdly, the traders use the foreign futures price as information in predicting future commodity prices. Like Kawai and Zilcha (1986), we nd that without the correlation between commodity prices and exchange rates, the traders optimal position in the forward exchange market is full hedging of his expected payment and receipt in foreign currency and the speculation in the domestic currency market. The optimal futures position is a partial hedge of spot position, speculation in the futures market and the eect of speculation in the forward market. We also nd that the separation theorem does not apply to the optimal spot position here. The spot position will be greater when the traders hedge their risks. Oshore hedging of all traders will be more eective when they hedge both price and currency risks. While increases in variances of prices and exchange rates decrease traders optimal spot and futures positions, a rise in the covariance between the domestic spot price and the foreign futures price in domestic currency can increase their optimal positions. The decreases in the dierence between the expected spot exchange rate and the forward exchange rate, the increase in exchange rate volatility, and the increase in correlations of exchange rates and domestic commodity prices can either increase or decrease the commodity spot price; this depends on which traders hedge their exchange rate risk and whether the eects of these changes on the supply is greater than the

26

eects on the demand in the markets. The empirical result also supports the theoretical result. We nd that Thai rice and rubber sheet prices are aected by exchange rates though exports and oshore hedging. Oshore hedging by dierent trader types leads to dierent impacts of foreign futures prices and foreign exchange rates on domestic commodity prices. With the biasedness of the forward foreign exchange market, we nd the eect of the forward foreign exchange rate on domestic rubber and rice prices even though rice prices (in domestic currency) and the exchange rate are not correlated. The model also indicates that some traders hedge their price risk in the foreign futures market, some do not hedge their risk, and some use the futures price as information to predict the future spot price. Further research could compare the optimal strategy, price volatility and traders welfare before and after the existence of domestic commodity futures market. The nding would guide the policy makers whether the country should have a domestic commodity futures market. Empirical work could be done using recent commodity prices to investigate how the prices and trader behaviour change when the domestic commodity future market becomes available. It could also apply the prices of more commodities in dierent countries to calculate the optimal hedging ratios based on the optimal strategy found in this paper.

References

[1] Anderson, R. W. and Danthine, J. P. (1983). Hedger diversity in futures market. The Economic Journal, vol. 93, 370-389. [2] Antoniou, A. (1986). Futures markets: Theory and tests. D.Phil Thesis, The University of York . [3] Bahmani-Oskooee, M. and Mitra, R. (2008). Exchange rate risk and commodity trade between the U.S. and India. Open Economic Review, vol. 19, 71-80. [4] Baillie, R.T. and Selover, D.D. (1987). Cointegration and models of exchange rate determination. International Journal of Forecasting, vol. 3, 43-51. [5] Benninga, S. and Oosterhof, C. (2004). Hedging with forwards and puts in complete and incomplete markets. Journal of Banking anf Finance, vol. 28, 1-17.

27

[6] Branson, W.H. (1969). The minimum covered interest dierential needed for international arbitrage activity. Journal of Political Economy, vol. 77, 1028-1035. [7] Chen, Y, Rogo, K. and Rossi, B. (2010). Can exchange rates forecast commodity prices? The Quarterly Journal of Economics, vol. 125 (3), 1145-1194. [8] Corbae, D. and Ouliaris, S. (1988). Cointegration and tests of purchasing power parity. Review of Economics and Statistics, vol. 55, 508-511. [9] Danthine, J. P. (1978). Information, futures prices, and stabilising speculation. Journal of Economic Theory, vol. 17, 79-98. [10] Dawe, D. (2002). The changing structure of the world rice market, 1950-2000. Food Policy, vol. 27, 355-370. [11] Deaton, A. and Laroque, G.(1992). On the behaviour of commodity prices. Review of Economic Studies, vol. 59, 1-23. [12] Deaves, R. and Krinsky, I. (1995). Do futures prices for commodities embody risk premiums? Journal of Futures Markets, vol. 15, 637-648. [13] Feder, G., Just, R.E., and Schmitz A. (1980). Futures markets and the theory of the producer under price uncertainty. Quarterly Journal of Economics, vol. 97, 317-328. [14] Gilbert, C.L. (1991). The response of primary commodity prices to exchange rate changes. In L.Philip (ed.), Commodity, Futures and Financial Markets: Advanced Studies in Theoretical and Applied Econometrics. Netherlands: Kluwer Academic Publishers, 87-124. [15] Glosten, L. R. and Milgrom, P.R. (1985). Bid, ask and transaction prices in a specialist market with heterogeneously informed traders. Journal of Financial Economics, vol. 14, 71-100. [16] Holthausen, D.M. (1978). Hedging and the competitive rm under price uncertainty. American Economic Review, vol. 69, 989-995. [17] Jin, H.J. and Koo, W.W. (2002). Oshore commodity and currency hedging strategy with hedging costs. Agribusiness and Applied Economics Report, vol. 483. 28

[18] Kawai, M. and Zilcha. I (1986). International trade with forward-futures markets under exchange rate and price uncertainty. Journal of International Economic, vol. 20, 83-98. [19] Kofman, P. and Viaene, J. M.(1991). Exchange rates and storable prices. In L.Philip (ed.), Commodity, Futures and Financial Markets: Advanced Studies in Theoretical and Applied Econometrics. Netherlands: Kluwer Academic Publishers, 125-152. [20] Sachsida, A., Ellery, R. Jr. and Teixeira, J. R. (2001). Uncovered interest rate parity and the Peso problem. Applied Economics Letters, vol. 3, 179-182. [21] Schmittmann, J.M. (2010). Currency heding for international portfolios. IMF Working Paper no. WP/10/151. [22] Timmer, C.P. (2009). Did speculation aect world rice prices? FAO ESA Working Paper no. 09-07. [23] Wakita, S. (2001). Eciency of the Dojima rice futures market in Tokugawa-period Japan. Journal of Banking and Finance, vol. 25, 535-554. [24] Wang, P. (2000). Testing PPP for Asian economics during the recent oating period. Applied Economics Letters, vol. 7, 545-548. [25] Yun, W. and Kim, H.J. (2010). Hedging strategy for crude oil trading and the factors inuencing hedging eectiveness. Energy Policy, vol. 38, 2404-2408.

f 2 = Yit t Cit Sit Sit + {Mit+1 (f (Sit ) + it+1 ) f f t+1 + (et+1 ft )Zit }. Yit f f and Zit at time by maximising his expected utility He choose Sit , Yit

(29)

Vi =

f Yit ,

f Yit ,Zit

(30)

29

f Eit ( it+1 ) = Eit (Mt+1 )f (Sit ) + Eit (Mit+1 it+1 ) Yit Eit (t+1 ) f + [Eit (et+1 ) ft ] Zit

f2 2 V arit ( it+1 ) = V arit (Mit+1 )f (Sit ) + V arit (Mit+1 it+1 ) + V arit (t+1 )Yit f2 +Zit V arit (et+1 ) + 2f (Sit )Covit (Mt+1 , Mt+1 it+1 )

f f Covit (Mit+1 it+1 , t+1 ) + 2Zit Covit (Mit+1 it+1 , et+1 ) 2Yit f f Zit Covit (t+1 , et+1 ) 2Yit

where Covit ( it+1 , et+1 ), Covit (Mit+1 it+1 , t+1 ) and Covit (Mit+1 it+1 , et+1 ) are equal to 0. His FOC with respect to Sit is Vi Xst = Ct i Sit + [Eit (Mit+1 ) Ai fS {f (Sit )V arit (Mit+1 ) + Covit (Mit+1 it+1 , Mit+1 )

f f Covit (Mit+1 , t+1 ) + Zit Covit (Mit+1 , et+1 )}] = 0 Yit

(31)

where fS =

f (Sit ) Sit .

Vi Yfft

f = t Eit (t+1 ) Ai {Yit V arit (t+1 ) Eit (f (Sit ))Covit (Mit+1 , t+1 ) f Covit (t+1 , et+1 )} Zit

= 0 and Vi

f Zit f f = Eit (et+1 ) ft Ai {Zit V arit (et+1 ) Yit Covit (t+1 , et+1 ) + Sit Covit (Mit+1 , et+1 )}

(32)

= 0. Appendix B The eect of changes in price and exchange rate volatilities are as follows.

Sit 2it pit <0 = Sit V arit (Mit+1 ) 3 (M 2it V arit it+1 )

(33)

30

Corrit (Mit+1 , t+1 ) Corrit (t+1 , et+1 )Corrit (Mit+1 , et+1 ) 2 ( 1 Corrit t+1 , et+1 ) q it (Mit+1 ) (Eit (et+1 ) ft ) VVar 3 ar (et+1 )

it

> 0

Sit

V arit (et+1 )

(Corrit (t+1 , et+1 )Corrit (Mit+1 , t+1 ) Corrit (Mit+1 , et+1 )) < 0

Sit Corrit (Mit+1 , et+1 )

(Corrit (t+1 , et+1 )Corrit (Mit+1 , t+1 ) Corrit (Mit+1 , et+1 ))2 } 2 ( 1 Corrit t+1 , et+1 ) < 0

Sit

q V arit (Mit+1 ) V arit (t+1 ) (t Eit (t+1 ) + (Eit (et+1 ) ft )Corrit (t+1 , et+1 ))

2 ( it (1 Corrit t+1 , et+1 ))

(Corrit (Mit+1 , t+1 ) Corrit (Mit+1 , et+1 )Corrit (t+1 , et+1 )) 2 ( it (1 Corrit t+1 , et+1 )) > 0 Appendix C V ar(AB ) = E [AB E (AB )]

2

(34)

where A and B are random variables. Dene X = A E (A) and Y = B E (B ). Then, A = X + E (A) and B = Y + E (B ). Therefore, V ar(AB ) = E [XY + XE (B ) + Y E (A) Cov (A, B )]2 = E [X 2 Y 2 + 2X 2 Y E (B ) + 2XY 2 E (A) 2XY Cov (A, B ) + X 2 E (B )2 + Y 2 E (A)2 +2XY E (A)E (B ) 2XE (B )Cov (A, B ) 2Y E (A)Cov (A, B ) + Cov 2 (A, B )] As E (X 2 Y 2 ) = V ar(A)V ar(B ) + 2Cov 2 (A, B ), E (X 2 Y E (B )) = 0, E (XY 2 E (A)) = 0, 31 (35)

V ar(AB ) = V ar(A)V ar(B ) + Cov 2 (A, B ) + E (B )2 V ar(A) +E (A)2 V ar(B ) + 2Cov (A, B )E (A)E (B ) (36)

V ar(Lt+1 ) = V ar(

Lt+1 Lt+1 Lt+1 )V ar(et+1 ) + Cov 2 ( , et+1 ) + E (et+1 )2 V ar( ) et+1 et+1 et+1 Lt+1 Lt+1 Lt+1 2 ) V ar(et+1 ) + 2Cov ( , et+1 )E ( )E (et+1 ) +E ( et+1 et+1 et+1

(37)

V ar(Lt+1 ) = V ar(

Lt+1 Lt+1 Lt+1 2 )V ar(et+1 ) + E (et+1 )2 V ar( ) + E( ) V ar(et+1 ) et+1 et+1 et+1

(38)

32

- Sysco Weekly Market Trends Report 11/9/2012Hochgeladen vonBuyEfficient
- Solutions Chapters 10 & 11 Transactions and Economic ExposureHochgeladen vonrakhiitsme
- ADFINAPHochgeladen vonKenneth Bryan Tegio
- midterm.docHochgeladen vonalellie
- Assignment # 1, International FinanceHochgeladen vonabdulhayee
- GS Gold BuyRec CommentHochgeladen vonadmin3341
- A Study on Financial Derivatives (Futures & Options)Hochgeladen vonVinod Ambolkar
- Sugar Jan 1 2010Hochgeladen vonAndysTechnicals
- RDHochgeladen vonDrYogesh Mehta
- portfolioHochgeladen vonSatinder Shinda
- Euro Money Article 101109Hochgeladen vonNaveen K. Jindal
- Indian Commodity Futures MarketsHochgeladen vonMahesh Jadhav
- Inrernationa Monetary FundHochgeladen vonAjit Saini
- Forex MgmtHochgeladen vonAnwar Siddique
- BasicsHochgeladen vontriptim_3
- Top50_2010Hochgeladen vongoldtrader
- MCom06Hochgeladen vonNidhi Singh
- Futures and Options Trading for Hedge Funds. the Regulatory Article)Hochgeladen von97989
- Derivatives v 2Hochgeladen vonPooja Bansal
- 43747635Hochgeladen vonKunal Chopra
- A Study on Trading Mechanisms of Commodities Market in IndiaHochgeladen vonKrishna Prince M M
- 1239286220863 Craig Baker Commodity Price Risk Management (1)Hochgeladen vonssinh
- Global Derivatives - Products, Theory and Practices -Chap01Hochgeladen vonUmlol Titani
- Qsl Update 20 6 14_cskHochgeladen vonCarla Keith
- 3013238Hochgeladen vonGauriGan
- 3012138Hochgeladen vonGauriGan
- Investor Presentation [Company Update]Hochgeladen vonShyam Sunder
- Derivatives 1Hochgeladen vonValentin Is
- FINA3203 02 FuturesHochgeladen vonBernice Chan Wai Wun
- Ch01HullOFOD9thEdition_-_Edited.pptxHochgeladen vonHarshvardhan Mohata

- VAR_3Hochgeladen vonPiyush Shah
- MEC-004 - Eco of Growth n Devlpt 2015Hochgeladen vonAshwin R John
- MEC-003 QT 2014Hochgeladen vonAshwin R John
- MEC-002 - Macro Analysis 2014Hochgeladen vonAshwin R John
- ANALYSIS OF FIN STATEMENTHochgeladen vonPrithviraj Kumar
- MEC-001 - Micronomic Analysis 2015Hochgeladen vonAshwin R John
- IGNOU MA ECO ASSIGNEMNTS JULY 2015 SESSIONHochgeladen vonAnonymous IL3nivBc
- 01. Theory of Consumer Behaviour - Basic ThemesHochgeladen vonRanjeet Kumar
- CMA FAQsHochgeladen vonAshwin R John
- Equivalence Between DiscountedHochgeladen vonAshwin R John
- The Correct Definition for the Cash Flows to Value a FirmHochgeladen vonAshwin R John
- Company Valuation MethodsHochgeladen vonKlaus Laur
- How the Subprime Crisis Went GlobalHochgeladen vonAshwin R John
- Kapil Gupta Balwinder Saini Submission 29Hochgeladen vonnishant_mendiratta_1
- RBI MSE Joint Initiative on Modelling the Indian EconomyHochgeladen vonAshwin R John
- Relationship Between Gold Price and SensexHochgeladen vonAshwin R John
- Momentum Strategies in Futures Markets and Trend-following FundsHochgeladen vonAshwin R John
- Gold Prices and Financial Stability in IndiaHochgeladen vonAshwin R John
- SSRN-id1487905Hochgeladen vonAshwin R John
- Identification of Bubble in Gold and Study of Its Propagation to Equity MarketsHochgeladen vonAshwin R John
- Swap Rate PaperHochgeladen vonAshwin R John
- Axe of Creation Am Scale RiffHochgeladen vonAshwin R John
- VAR Limitations in Practice Feb03Hochgeladen vonAshwin R John
- Becker CPA 2011 Audit 5Hochgeladen vondonkalonk
- 2011 a-1A Summary NotesHochgeladen vonAshwin R John
- J 17 13 III (Management)Hochgeladen vonAshwin R John

- SMEs in Msia_ebook ISBN 978-3-659-52051-8Hochgeladen vonPok Wei Fong
- Balance of Payments( COMPLETE NOTES )Hochgeladen vonaisharafiq1987
- 10 Profile of Investment IncentivesHochgeladen vonAnonymous zp7PyaV7
- Ppt 8 International Monetary System.pdfHochgeladen vondevesh_mendiratta_61
- indonesia-2016-OECD-economic-survey-overview-english.pdfHochgeladen vonFabian Dimas
- ficotroubleshooting-131122184332-phpapp01Hochgeladen vonv_k84
- Project Report on MultiCurrency Forex Card of Thomas CookHochgeladen vonarjun lal
- Chapter 12 _Class PresentationHochgeladen vonCecilia Ooi Shu Qing
- larsen10_ch12Hochgeladen vonabmyonis
- Japan Early 30s PaperHochgeladen vonJohn Brimlow
- International Finance and Trade [Emba]5207 2011 EditionHochgeladen vonSam Madzokere
- Gundlach Just MarketsHochgeladen vonZerohedge
- milk-newHochgeladen vonSyam Sundar
- L&W - Field Mapping(2006-03-30)3.0Hochgeladen vontriton45
- Asset Market Approach : PakistanHochgeladen vonMuhammad Rehan Tahir
- Financial Stability ReporHochgeladen voncaport
- Unit-11 International Pricing PolicyHochgeladen vonbhar4tp
- MONETARY HISTORY OF BANGLADESH: 1972 - 2011Hochgeladen vonSharjil Haque
- 2009 ASEA YearbookHochgeladen vonAfricanExchanges
- Financial IHochgeladen vonanggiirawanfawfaw
- BIS - Roles and Objectives of Central BanksHochgeladen vondascXC
- Upload Exchange Rate From ECBHochgeladen vonpawans8110
- RBI Intervention in Foreign Exchange MarketHochgeladen vonshamchandak03
- What is Exchange RateHochgeladen vonMELAKU
- Functions of Foreign Exchange MarketsHochgeladen vontechybyte
- International Finance MultipleHochgeladen vonVăn Dương
- Macro FinalHochgeladen vonpitimay
- Canada's Currency Regime: Options and AnalysisHochgeladen vonSean McConnachie
- Document (1)Hochgeladen vonVijay
- S&P affirms SA's junk credit ratingsHochgeladen vonFin24

## Viel mehr als nur Dokumente.

Entdecken, was Scribd alles zu bieten hat, inklusive Bücher und Hörbücher von großen Verlagen.

Jederzeit kündbar.