Intend To Examine The Causal Relationship

Published Date: 02 Nov 2017

1.0 Introduction

The Asian financial crisis of 1997-98 raises attention towards the dynamic linkage between stock prices and exchange rates. During the financial crisis period, economy of many Asian countries was plummeted as the continuous currency depreciation as well as stock markets was hit badly. The relationship between stock prices and exchange rates again become hot topic among international investors who involved in large cross-border movement of funds due to portfolio investment as their profitability are always exposed to currency risk.

Besides, linkage between the exchange rates and stock prices has also received a high attention of researchers and policy makers in the last few decades. With the emergence of new capital markets, the adoption of more flexible exchange rate policies during the generalized floating of the major currencies in early 1973, and the relaxation of foreign capital controls have open up a wide space for researchers to broaden the literature in this topic. The way where exchange rates influence stock prices has never been so important before. The research of this area has mainly focused on the causality determination between stock prices and exchange rates in developed countries and developing countries, for us, it will be developing countries.

The current state of research cannot provide us with either theoretical or empirical consensus on the topic; however, we wish to contribute by further investigating between two South East Asia countries, name as Malaysia and Singapore.

Research Background

In our study, we intend to examine the causal relationship between stock market indexes and exchange rates among two South East Asia countries, which are Malaysia and Singapore.

Economy is the backbone of one country, which sustains the governance policy in a country such as administration, judicature and military affairs. Each country has its own economic history; at here we briefly described the economic history of Malaysia and Singapore separately.

1.1.1 Malaysia Economy

Malaysia is a multiethnic, upper-middle income nation that relied heavily on income from its natural resources such as palm oil, petroleum and natural gas to engineer successful diversification into manufacturing and rise dramatically for all racial groups (Yusof & Bhattasali, 2008). It has continued to enjoy relative prosperity, and as a commodity exporter with total income increasing at 6 to 7 percent each year from 1970 until 2000.

In 1995, the GDP peaked at 44 percent. However, it had plummeted after the Asian financial crisis 1997. In mid-May 1997, the Thai baht came under severe pressure from speculative attacks. The ringgit was also not spared, and came under severe selling pressure. Bank Negara Malaysia’s (the central bank of Malaysia) immediate response was to intervene in the foreign exchange market to uphold the value of the ringgit. Bank Negara valiantly upheld the value of the ringgit for about a week before it finally was forced to float the ringgit on July 14. By that time, the bank had already lost close to U.S. $1.5 billion in the effort to prop up the ringgit. At its lowest point, the ringgit depreciated against the dollar by almost 50 per cent, hitting a high of RM 4.88 to the U.S. dollar on January 7, 1998.

After a brief period of stability during February and March, the exchange rate continued to deteriorate with wide fluctuations in the following months. Even more drastic than the plunge in the exchange rate, was the collapse of the stock market. Between July and December 1997, the composite index of the Kuala Lumpur Stock Exchange (KLSE CI) fell by 44.9 per cent. Following a slight recovery in the first quarter of 1999, the index again fell, this time to an eleven-year low of 262.70 points on September 1, 1998. On the whole, between July 1, 1997 and September 1, 1998, market capitalization in the KLSE fell by about 76 per cent to RM 181.5 billion. In fact, although it enjoyed the best pre-crisis economic fundamentals among countries that were hit by the crisis, Malaysia experienced the biggest stock market plunge in the region.

It had stood at an average 22 percent of GDP since 2000 and ran abruptly into the wall of the 2001 worldwide slowdown. Globally, foreign direct investment dropped approximately 50 percent and in Malaysia the decrease was an even more precipitous 85 percent. GDP growth dropped to 0.7 percent for 2001, from its normal 7 percent to 9 percent. With a combination of internal and external factors dampened growth over the period 2001 to 2007 to around 4 percent per year.

During the year of 2001 and 2002, Malaysia was hard hit by the global economic downturn and the slump in the information technology (IT) sector. The GDP in 2001 raise only 0.5 percent due to estimated 11 percent contraction in exports, but an important fiscal incentive package equal to US $1.9 billion mitigated the worst of the recession and the economy rebounded in 2002 with a 4.1 percent increase. In the half year of 2003, the economy grew 4.9 percent, however when external pressures from SARS and the Iraq war led to prudence in the business community.

Later, real growth rate of Malaysia GDP in 2008 has been found to be around 5.5 percent. Various types of sectors had contributed individually for Malaysia economic development. In financial year 2008, industrial sector and service sector had largest contribution which is 44.6 percent and 45.7 percent respectively. At end of 2012, Malaysia’s capital market crossed the RM 2 trillion thresholds for the first time. The capital market had achieved an annual compounded growth rate of 11 percent which there is rapid industry expansion and strong regulatory oversight that underpinned investor in the Malaysia capital market.

Based on historical data, the KLCI averaged 690.4 reaching an all time high of 1606.6 in April 2012 and a low of 89 in April of 1977 from the period 1977 to 2012. KLCI is a major stock market index which tracks the performance of large companies based in Malaysia.

1.1.2 Singapore economy

Singapore has a highly developed market-based economy. Singapore is one of the Four Asian Tiger which along with Hong Kong, South Korea, and Taiwan which characterized by an export-oriented economy, relatively equitable income distribution, trade surpluses with United States and other developed countries.

The years of the comparatively untroubled 1980s were followed by the Asian Financial Crisis of 1997. The value of exports from Singapore had grown by 15 percent per year over the period of 1990 to 1996. On 17th July 1997, the Singapore Dollar (S$) dropped quickly in value from US$ 1 to S$ 1.495 prior to the depreciation to US$ 1 to S$ 2.68 a few days later. Overall, the economy of Singapore was contracted by 1.4 percent in 1998 in terms of real gross domestic product (GDP).

According to Tan (1999), the real effective exchange rate of the Singapore Dollar which adjusted downwards by approximately 8 percent from June 1997 had reflects the effect of trade weights. Besides, Singapore’s Stock Market, Straits Time Index (STI) also suffered severe devaluation of -47 percent from 9th July, 1997 until 5th August 1998. The downturn sharply had significant adverse effects with increase in non-performing loans, decrease in consumption and investment and, equally important, confidence.

Rajan and Siregar (2002) determined that the exchange regime of Singapore may be an effective measurement in order to maintain the stability of domestic price, and export competitiveness for small and open economies. Furthermore, it provides a large degree of flexibility when suffer great economic fluctuations. For instance, during the summit of the Asian Financial Crisis during 1997-1998, the Monetary Authority of Singapore (MAS) permitted for a 20 percent depreciation of the Singapore dollar in order to cushion and conduct the economy to a soft landing. It had stood Singapore in good stead during crisis. Yet, the panic of rising non-performing loans caused the banks to tighten credit, increasing interest rate spreads and decreasing loan-deposit ratios. As a result, local domestic demand was consequently dampened, and contributing to the recession.

The Singapore economy was showing recovering signs by early 1999. In the first quarter of 1999, the economy returned with a positive growth. The recovery was sustained through the year overall GDP for the year 1999 grown by 7.2 percent. The densely-populated country weathered the crisis more stolidly than other economies by working towards direct cost cutting measures, such as wage and operating cost reductions to sustain its competitiveness. The economy was picked up soon and registered a growth rate of 9.4% in 2000.

All those decades of hard work were almost ruined when the global financial crisis such as Asian financial crisis 1997, United States subprime mortgage crisis 2007, and European sovereign debt crisis 2009 hit the Singaporean economy badly. The country has always relied on exports to fund its economic growth, and as demand slumped in key markets in the U.S., and the Europe, its export industry took a beating. The Singapore economy contracted 2 percent with the construction sector the only one to post double-digit growth in 2009. Manufacturing industry, the biggest share of the country’s GDP contracted by 4.1 percent. Yet, the Singaporean economy picked up as the global economy did so.

Singapore’s economy has been ranked amongst the world’s ten most open, competitive and innovative. It also rated as the most business-friendly economy in the world and is a hub of many multinational companies. Straits Time Index (STI) is a major stock market index based in Singapore, and from 1999 until 2012, the STI averaged 2318.5 reaching an all time high of 3831.2 in October 2007 and low record of 1170.9 in March 2003 based on the historical data.

1.2 Problem Statement

There are several reasons that motivate us to conduct the study. First, we argued that after an economic crisis such as the Asian Financial Crisis in 1997, there may cause the changes in nature of relationship between stock prices and exchange rates (Pan, Robert & Liu, 2007). Meanwhile, it has been never exposure all variables that significantly affect in the stock market (Kim, 2003). In addition, Pan et al. (2007) supported that the depreciation in currency values and the crashing down of stock prices during the crisis reinforce the conventional impression that exchange rates and stock prices tend to move in a tandem. Take a look in Malaysia, the traditionally stable ringgit collapsed and it was followed by the collapse of the KLSE whereby Kuala Lumpur Composite Index dropped by 44.8% during the financial crisis. However, Paul, Theodore and Xu (2011) stated that neither theoretical nor empirical studies has accomplish consensus on the relationship between these two variables.

Furthermore, empirical studies have not disclosure a standard model that sufficiently deals with the non-stationary of the variables (Baharom, Royfaizal & Habibullah, 2008). In order to avoid the results of spurious regression, Baharom et al. (2008) supported that treating non-stationary variables is necessary before the hypothetic testing. Moreover, the possible structural breaks in the model are rare investigated by previous researchers (Kim, 2003 and Lean, Narayan & Smyth, 2011). That is, there is a great potential in order to use a mixture of developed and developing countries as we argue that the variables causality may be greater in countries with developed foreign exchange and stock market (Kim, 2003).

Finally, based on our knowledge, it is few studies that focus on developing countries such as Malaysia and also found conflicting results which encourages us to conduct the study. Previous researchers are put more attention in developed countries like the United State (Pan et al., 2007). Due to the effect on emerging capital market, for instance the Malaysia, there is an increasing in the demands to study the relationship between stock prices and exchange rates in such developing countries.

Research Objectives

General Objective

Our main objective is to investigate whether there is a relationship between stock prices and exchange rates or vice versa and whether the crisis adversely changed the causal interaction between exchange rates and stock prices between 2 South East Asia, Malaysia and Singapore.

Specific Objectives

The study determined specific objectives as below:

To investigate the relationship between stock prices and exchange rates or vice versa among Malaysia and Singapore.

To determine whether the crisis adversely changed the causal interaction between exchange rates and stock prices among Malaysia and Singapore.

To determine whether long-run or short-run relationship exists between stock prices and exchange rates.

Research Questions

Is there any relationship between stock prices and exchange rates or adversely among Malaysia and Singapore?

Are there any changes in the relationship between stock prices and exchange rates among Malaysia and Singapore during or after Asian financial crisis?

Is there exists of long-run or short-run relationship between stock prices and exchange rates among Malaysia and Singapore?

Hypotheses of the Study (for quantitative research)

The hypotheses of this research project are developed as below:

Significant of the relationship:

H0: There is no significant relationship between stock prices and exchange rates (t = 0)

H1: There is a significant relationship between stock prices and exchange rates (t ≠ 0)

Stationarity:

H0: Stock price is non-stationarity (Stock price has unit root), 

H1: Stock price is stationarity (Stock price has no unit root), 

H0: Exchange rate is non-stationarity (Stock price has unit root), 

H1: Exchange rate is stationarity (Stock price has no unit root), 

Short-run and Long-run relationship:

H0: There is no cointegrating vector between exchange rates and stock prices (r=0).

H1: There is cointegrating vector between exchange rates and stock prices (r ≠ 0).

Granger causality relationship:

H0: Stock prices do not Granger cause exchange rates during Asian financial crisis among Malaysia and Singapore.

H1: Stock prices do Granger cause exchange rates during Asian financial crisis among Malaysia and Singapore.

H0: Stock prices do not Granger cause exchange rates after Asian financial crisis among Malaysia and Singapore.

H1: Stock prices do Granger cause exchange rates after Asian financial crisis among Malaysia and Singapore.

H0: Exchange rates do not Granger cause stock prices during Asian financial crisis among Malaysia and Singapore.

H1: Exchange rates do Granger cause stock prices during Asian financial crisis among Malaysia and Singapore.

H0: Exchange rates do not Granger cause stock prices after Asian financial crisis among Malaysia and Singapore.

H1: Exchange rates do Granger cause stock prices after Asian financial crisis among Malaysia and Singapore

Significance of the Study

Establishing the relationship between stock prices and exchange rates is important as the following reasons:

To provide further evidence on the relationship between exchange rates and stock prices among Malaysia and Singapore during year 1997 to year 2012.

To broaden the existing literature on the relationship between exchange rates and stock prices in Malaysia and Singapore during year 1997 to year 2012.

The link between the two markets may be used to predict the path of the exchange rate, in order benefit multinational corporations in managing their exposure to foreign contracts and exchange rate risk stabilizing their earnings;

Currency is more often being included as an asset in investment funds' portfolios. Knowledge about the relationship between currency rates and other assets in a portfolio is vital for the performance of the fund;

The understanding of the relationship between stock prices and exchange rates may prove helpful to foresee a financial crisis.

1.7 Chapter Layout

The remainder of the paper is organized as follows. Chapter 2 provides literature review. Chapter 3 discusses the data, time frame considered right through the study and methodological issues. Chapter 4 present empirical results and findings. The discussion, conclusion and implications are given in Chapter 5.

1.8. Conclusion

Chapter 1 provides an overall economic historical background of the two South East Asia which are Malaysia and Singapore. It provides an overall understanding on the topic and purposes of conducting the research. Our study mainly focuses on the causality relationship between stock prices and exchange rates among Malaysia and Singapore during and after Asian financial crisis 1997. In order to more understanding about the topic, a review on relevant literature has to be done to seek for supporting evidence for the study. This will be conducted in the Chapter 2.

CHAPTER 2: LITERATURE REVIEW

2.0 Introduction

In this chapter, a review of the literature will be discuss on the topic that study in Chapter 1, the relationship between stock prices and exchange rates in two South East Asia which are Malaysia and Singapore. For further research, a relevant theoretical model is adopted to develop the proposed conceptual framework and the proposed conceptual framework is developed based on the research objectives and research questions that mentioned in Chapter 1. Besides, hypotheses are developed to further explain the proposed conceptual framework.

2.1 Relevant Theoretical Models in Studies

From the past studies, the relationship between exchange rates and stock prices is far from certainty. There are two main models that relate these financial markets, which are (1) flow-oriented and (2) stock-oriented.

2.1.1 Flow-oriented model

According to Dornbusch and Fisher (1980), it suggest the "flow-oriented" models of exchange rates, which speculate that changes in exchange rates affect global competitiveness and trade balances, thereby affecting real profit and output. This theory mentioned that the depreciation of the local currency makes the local firms more competitive, leading to an increase in their export and consequently price of the stock will increase. The past researcher, Tabak (2009) mention that the transmission channel would be the volatility of the exchange rates which affect firm’s values via changes in competitiveness and changes in the value of firm’s assets and liabilities which dominated in foreign currency, at last, affecting firm’s net profit and thus the equity value. In other words, it states that exchange rate Granger cause stock prices (Lean, Narayan, & Smyth, 2011).

With the changes of exchange rate, it may affect the competitiveness of firms on input and output price. When the exchange rate appreciates, since exporters will shrink and the stock prices will decline, while importers will increase their competitiveness in local markets, and thus their profit and stock prices will increase. The depreciation of exchange rate will make an adverse effect on exporters and importers. Based on Yau and Nieh, (2006) study, they found that exporters will have advantage against other countries’ exporters and their sales as well as stock prices will be higher. Currency appreciation has brings both a negative and positive effect on domestic stock market for an export-dominant and an import-dominated country, respectively.

Furthermore, stock price can be affected by exchange rate not only for multinational and export oriented firms but also for domestic firms (Aydemir & Demirhan, 2009). For a multinational company, exchange rates change will result in both immediate changes its foreign operations’ value and a long-term change in profitability of its foreign operations reflected in successive income statements. Thus, the changes in economic value of firm’s foreign operations may influence stock prices. The local firms also can be affected as they may import a part of their inputs and export their outputs. For example, a depreciation of its currency makes imported inputs more expensive and exported outputs cheaper for a firm. Therefore, the depreciation has the positive effect for export firms (Aggarwal, 1981) and the income of the firm may also increase, consequently, the average level of stock price will keep increasing.

2.1.2 Stock-oriented model

Secondly, "stock-oriented" models emphasize the capital account as the major determinant of exchange rate. Branson (1983) and Frankel (1983) viewed that exchange rate as equating the supply and demand for assets such as stocks. This theory determines exchange rate dynamics by giving the capital account an important role. Since the values of financial assets are determined by the stock prices, expectations of relative currency values play a significant role in their price movements. Thus, stock price innovations may influence, or influenced by, exchange rate dynamics. For example, when the foreign currency (the US dollar) depreciates, it will increase returns on the foreign currency (US dollar). Such activities will motivate investors to move funds from local assets (stocks) towards dollar assets, gloomy stock prices. Therefore, a depreciation of currency brings a negative effect on stock market return.

There are two subset of the stock oriented which known as portfolio balance approach and monetary approach:

2.1.2.1 Portfolio balance model

Portfolio balance models hypothesize that when the stock prices move up, it will drive up the interest rate of local currency with the resulting effect of a fall in the exchange rate (Branson, 1983). Moreover, Markowitz (1952) had made a simple assumption where the risk is defined by volatility. This theory considers investors are risk adverse who willing to accept more volatility for higher payoffs and will accept lower returns for less risk investment. He also explained that investor need only invest in one security with maximum expected returns to maximize the expected value of a portfolio (Markowitz, 1991).

In other words, changes in stock prices will give impact on exchange rate movements (Aydemir & Demirhan, 2009). This theory stated that stock price is expected to lead exchange rate with a negative correlation since a decline in stock prices reduce the wealth of domestic, which will leads to lower domestic money demand and interest rates. In addition, the prices of local stock reduce leads foreign investors to lower demand for local assets and local currency. This shifts the demand and supply of currencies cause capital outflows and the depreciation of domestic currency. In contrast, when there is increase in stock price, the foreign investors become willing to invest in a country’s equity securities and get advantage from international diversification. This may lead to inflows of capital and an appreciation of currency (Caporale, Pittis, & Spagnolo, 2002; Stavárek, 2005; Pan, Fok, & Liu, 2007).

2.1.2.2 Monetary model

However, for monetary model (Gavin, 1989; Subayyal & Shah, 2011), it assumes that there is a weak or no linkage between stock prices and exchange rates except that both variables influenced by some common factors. In conclusion, these two models give us totally different theoretical results about the interaction and causality between exchange rate and stock price

2.2 Review of the Literature

2.2.1 Review of the stock prices

While stock represents the certificate of ownership in a company and stock prices should reflect business values of the respective companies, stock market trading activities by various participants could also have their influences in the fluctuations of stock prices.

Determining stock prices is a complex and conflicting task. Consistent with the economics theory, an asset price is usually determined by the supply and demand condition in a market (Pagano, 1988). Similarly in case of stock prices, it emerges by trading structure among the investors in stock markets.

Malik, Qureshi and Azeem (2012) stated that the major factors affect stock prices changes include the firm’s key performance indicators (fundamentals), market efficiency, investor’s perception, and some macroeconomic variables such as gross domestic products (GDP), inflation rates, interest rates and exchange rates.

First and foremost, empirical studies investigated that fundamental factors such as profitability, inflation, real interest rate et al., have an influence in stock prices changes (Mikkelson & Partch, 1985; Lakonishok & Shapiro, 1986; Pindyck, 1988; Barclay, Litzenberger & Warner, 1990; Abarbanell & Bushee, 1997; Domowitz, Glen & Madhavan, 1997; Morck, Yeung & Yu, 1999; Heaton & Lucas, 2000; Piotroski & Roulstone, 2004; Malik et al., 2012).

Furthermore, the most familiar interpretation for the large and unpredictable swings of common stock price is that price changes represent the efficient discounting of new information. Fama, Fisher, Jensen and Roll (1969), Grossman and Shiller (1981) and West (1988) stated that stock prices fluctuations has related to the efficiency market hypothesis (EMH).

On the other hand, LeRoy (1973), Jiambalvo, Rajgopal and Venkatachalam (2001) as well as Kurz, Jin and Motolese (2005) supported that investor’s perception will affect the performance of stock prices.

Last but not least, macroeconomic variables, for instance gross domestic products (GDP), money supply, gold reserve et al. have been observed to be significantly changed the volatility of stock prices (Cheung & Lai, 1999; Garcia, 1999; Brennan & Xia, 2001; Balke & Wohar, 2006; Alatiqi & Fazel, 2008; Baele, Bekaert & Inghelbrecht, 2009).

In our study, we only focus on one of the macroeconomic variable, names as foreign exchange rates, which will influence the stock price movements.

As the increasing international diversification, cross-market return correlations, gradual abolishment of capital inflow barriers and foreign exchange restrictions or the adoption of more flexible exchange rate arrangements in emerging and transition countries, stock market and foreign exchange market have become more interdependent. These changes have increased the variety of investment opportunities as well as the volatility of exchange rates and risk of investment decisions and portfolio diversification process (Aydemir & Demirhan, 2009).

Therefore, understanding the linkages between stock prices and exchange rates will help domestic or international investors for hedging and diversifying their portfolio as well as the policy holders governing in monetary policy.

2.2.2 Review of Exchange Rate

When there is the end of the Bretton Woods regime in 1973, the exchange rates have fluctuated widely in many countries (Joyce & Kamas, 2003). The exchanges rates have bring lots of vital effects on production, employment, and trade, thus it is important to understand the factors responsible for variations in exchange rates. There are many factors such as purchasing power parity (PPP), size of the country, current account, trade openness, productivity differential, interest rate differential, and term of trade, have an impact on the fluctuations of real exchange rate.

PPP or the relative prices of traded goods and services between countries is one of the famous variables to determine the exchange rate movement over a long time span. Meanwhile, most empirical evidence is unable to provide a satisfactory explanation for the exchange rate behavior in the short run, and generally believed to be associated with economic fundamentals in the long run (MacDonald, 1999). According to the researchers, Detken, Dieppe, Henry, Smets, and Marin, (2002); Chiu, (2008), the PPP absolute version states that the nominal exchange rate of two currencies should be proportional to a ratio of domestic and foreign prices in the long run. In MacDonald (1999), and Chiu (2008), they found that PPP is a useful device to capture the long run exchange rate behavior.

Besides, the past researchers also had add other monetary variables such as degree of openness, degree of international financial intervention, differential of inflation rates, differential of interest rate, differential of productivity, Gross Domestic Product (GDP), term of trade, interest rate, foreign reserves, net foreign assets, ratio of fiscal balance, (Maeso-Fernandez, Osbat, Schnatz, 2002; Joyce & Kamas, 2003; Zhang, & Pan, 2004; MacDonald, & Ricci, 2004; Candelon, Kool, Raabe, & Veen, 2006; Chiu, 2008; Ganguly & Breuer, 2010; Saeed, Awan, Sial, & Sher, 2012). They had found that the monetary variables are significance to the exchange rate volatility. The variables will affect the appreciation and depreciation of the exchange rate.

2.2.3 Studies of the relationship between stock prices and exchange rates

The influence of exchange rate fluctuation on the economy has become more crucial topic with the passage of time, especially after the Asian Financial Crisis 1997. Adoption of flexible exchange rate, financial and market integration, and the globalization have made the economy difficult to keep itself away from the influence caused by the volatility of exchange rates (Aurangzeb & Asif, 2012).

The concepts of whether the stock price and exchange rate are interrelated have been studied since 1970’s. Based on our knowledge, Franck and Young (1972) was the first study that inspected the relationship between these two variables. They use six exchange rates to study their volatilities and found that stock price does no have the significant interaction with the realignments of exchange rate.

Empirical studies dispute the correlation between stock prices and exchange rates has been started few decades ago. Since there is a various number of empirical research so far have been conducted to look into the correlation between these two variables, the researchers have found contradictory results regarding the existence and direction of their relationship, which has made the area confused environs of finance literature (Aurangzeb & Asif, 2012).

Among the few studies on emerging markets includes; Aggarwal (1981); Jorion (1990); He and Ng (1998); Dimitrova (2005); Phylaktis and Ravazzolo (2005); Richards, Simpson and Evans (2009); Yau and Nieh (2009); Tian and Ma (2010); as well as Lee, Doong and Chou (2011) found a significant positive relationship between stock prices and exchange rates. Conversely, some of the researchers counter above argument and presented a significant negative relationship between two variables (Kim, 2003; Adjasi, Harvey & Agyapong, 2008; Aydemir & Demirhan, 2009; Narayan, 2009). However, several studies like Nath and Samanta (2003), Ismail and Isa (2006), Rahman and Uddin (2009), Lean, and Narayan and Smyth (2011) have fail to find the correlation between stock prices and exchange rates on emerging markets.

Yang, Kolari and Min (2003) studied found out that both long-run and short-run co-integration relationships were strengthened during the crisis. Nonetheless, there are some researchers fail to investigate long-run relationship between stock prices and exchange rates, for instance the Muhammad and Rasheed (2002); Smyth and Nandha (2003); Tabak (2006); Baharom, Royfaizal and Habibullah (2008); Alhayky and Houdou (2009); Kumar (2009); Zhao (2010); and also Alagidede, Panagiotidis and Zhang (2011).

On the direction of causation, Hatemi and Roca (2005), and Pan, Fok and Liu (2007) have the mixed results. Whereas, several researchers have observed a bidirectional correlation between stock prices and exchange rates, for examples the Azman-Saini, Habibullah and Azali (2003); Hussain and Liew (2004); Azman-Saini, Habibullah, Law and Dayang-Afizzah (2006); Obben, Pech and Shakur (2006); Subayyal and Shah (2011); and also Rjoub (2012). On the other hand, Muhammad and Rasheed (2002) in addition Abdalla and Murinde (2010) found out a unidirectional causality from exchange rate to stock prices in the emerging countries.

Moreover, more researchers have a result support to portfolio model rather than the traditional flow-oriented model (Obben, Pech & Shakur, 2006; Tabak, 2006; Pan, Fok & Liu, 2007; Richards, Simpson & Evans, 2009; Ooi, Wafa, Lajuni & Ghazali, 2009; Abdalla & Murinde, 2010; Alagidede, Panagiotidis & Zhang, 2011). There are only few researches, for instance Tabak (2006) in addition Yau and Nieh (2009), which supported the traditional model.

Since there is neither theoretical nor empirical studies accomplish consensus results regarding the relationship and the direction causality between stock prices and exchange rates, which encourages us to conduct the relevant study in order to provide further evidence toward Malaysia and Singapore during and after the Asian financial crisis 1997.

2.3 Proposed Conceptual Framework

In this study, variables included are stock prices and exchange rate. Both of the variables, are acting as independent variables and independent variables at the same times, which mean, when stock price is independent variables, exchange rate will be the dependent variables we interested to study and vice versa.

On the theoretical side, there is no consensus towards the causality direction between stock prices and exchange rates. For instance, portfolio balance models of exchange rate determination postulate a negative relationship between stock prices and exchange rates and that the causation runs from stock prices to exchange rates.

In portfolio balance model, if individuals hold domestic and foreign assets, including currencies, in their portfolio and exchange rates play the role of balancing the demand for and supply of assets. An increase in domestic stock prices lead individuals to demand more domestic assets. However, in order to buy more domestic assets local investors would sell foreign assets and causing domestic currency appreciates. Events of increase of domestic assets price also lead investors to increase their demand for money, which in turn raises domestic interest rates. This again leads to appreciation of domestic currency by attracting foreign capital. Another channel for the same negative relationship is increase in foreign demand for domestic assets due to stock price increase. This would also cause a domestic currency appreciation.

In contrast to portfolio balance model, a positive relationship between stock prices and exchange rates with direction of causation running from exchange rates to stock prices can be explained when domestic currency depreciation makes local firms more competitive, leading to an increase in their exports. This in turn raises their stock prices.

Besides two models proposed by our previous researchers, Mookerjee and Yu uses efficient market hypothesis to explained the nature between stock prices and exchange rates. Efficient market hypothesis suggests that markets behave in an unpredictable manner in tandem with the inflow of news which is random in nature. Thus, stock market and foreign exchange markets should show only short-term association. If these two are cointegrated over a longer period of time, this would suggest that they are predictable (i.e. movement in one market can be used to predict the other market, rendering the efficient market hypothesis invalid. (Mookerjee and Yu, 1997)

Wu (2000) explains the positive and negative relationship between exchange rate and stock prices by a real interest rate and an inflationary disturbance. According to real interest rate disturbance, when the real interest rate rises, capital inflow increases and the exchange rate fall. However, since higher real interest rate reduces the present value of future cash flows, stock prices will decline. An inflationary disturbance may explain negative relationship between exchange rate and stock price. That is, when inflation increases, the exchange rate rises and because of high inflation expectations, investors will demand a higher risk premium and high rate of return. As a result, stock prices will decrease.

2.4 Hypotheses Development

Due to inconclusive and ambiguous causality direction, nature and relationship between stock prices and exchange rate, we could not suggest any form of hypotheses. We will stick to the hypotheses developed in Chapter 1 which is:

Significant of the relationship:

H0: There is no significant relationship between stock prices and exchange rates (t = 0)

H1: There is a significant relationship between stock prices and exchange rates (t ≠ 0)

Stationarity:

H0: Stock price is non-stationarity (Stock price has unit root), 

H1: Stock price is stationarity (Stock price has no unit root), 

H0: Exchange rate is non-stationarity (Stock price has unit root), 

H1: Exchange rate is stationarity (Stock price has no unit root), 

Short-run and Long-run relationship:

H0: There is no cointegrating vector between exchange rates and stock prices (r=0).

H1: There is cointegrating vector between exchange rates and stock prices (r ≠ 0).

Granger causality relationship:

H0: Stock prices do not Granger cause exchange rates during Asian financial crisis among Malaysia and Singapore.

H1: Stock prices do Granger cause exchange rates during Asian financial crisis among Malaysia and Singapore.

H0: Stock prices do not Granger cause exchange rates after Asian financial crisis among Malaysia and Singapore.

H1: Stock prices do Granger cause exchange rates after Asian financial crisis among Malaysia and Singapore.

H0: Exchange rates do not Granger cause stock prices during Asian financial crisis among Malaysia and Singapore.

H1: Exchange rates do Granger cause stock prices during Asian financial crisis among Malaysia and Singapore.

H0: Exchange rates do not Granger cause stock prices after Asian financial crisis among Malaysia and Singapore.

H1: Exchange rates do Granger cause stock prices after Asian financial crisis among Malaysia and Singapore

2.5 Conclusion

Chapter 2 has been started with theoretical foundation provided by past researchers which consist of three models, flow-oriented model, portfolio balance model and monetary model. Next we review all previous literature that relate to our research topic, discover the methodology and provide some findings in previous studies. With the foundation of theoretical framework and literature review, we then step into proposed theoretical frameworks which seek for the nature and the rationale behind the nature of the behavior between stock prices and exchange rates. We will stick to the original hypotheses that setup in Chapter 1 since there are no consensuses in both past theoretical framework and proposed theoretical framework. We will try to seek for answer in our studies and continue in Chapter 3 by starting to discuss methodology we use and data sources.

Chapter 3

3.0 Introduction

Chapter 3 will discuss about the methodology that used in our research. In detail, the description on the research design of quantitative research, data collection method of secondary data which include target population and sampling size are being discussed at here. Besides, data processing which includes checking, editing, coding and transcribing is explained at here too. Lastly, data analysis is carried out in this chapter.

3.1 Research Design

The research design refers generally to the plan that sketch to assimilate all components of the study in a logical and rational way, thereby, ensuring effective presentation of the research problem; it constitutes the layout of introduction of issue, literature review, analysis of data, data presentation and conclusion. (???)

3.2 Data Collection Methods

Based on our knowledge, previous researchers study in the dynamic between stock prices and exchange rates movements were examined in several Asian countries but only a few studied conducted between Malaysia and Singapore. Thus, our research paper focuses towards the associations between these two variables among Malaysia and Singapore.

In this project paper, secondary data of stock prices and exchange rates are retrieved from Thomson Reuters Datastream. The sources of data obtained through Thomson Reuters Datastream are from Kuala Lumpur Stock Exchange (KLSE)’s Main Board and FTSE Straits Times Index (STI) for stock prices independently, as well as from Bank Negara Malaysia (BNM) and Monetary Authority of Singapore for exchange rates respectively.

This study employed daily time series data of stock prices and exchange rates for two South East Asian countries, Malaysia and Singapore, spanning from Jun 1997 to December 2007, with a total number of 2590 daily observations for a country for each variable available in this analysis. The reasons for this research project to employ daily data as we try to: (1) avoid wide fluctuation in stock prices and exchange rates; (2) reduce the volatilities between stock prices and exchange rates; and (3) capture the sensitive movements between these two variables. In addition, aims to investigate the impact of Asian Financial Crisis 1997 on the relationship of the variables on these two countries, sample period that is divided into two sub periods that covers from 2nd Jun 1997 to 26th January 1998 (during crisis), and 3rd February 1998 until 31th December 2007 (post-crisis).

This project paper differs from other studies that we examine the impact between Malaysia and Singapore only by using two currencies involve direct quote of Malaysian ringgit (MYR) and Singapore Dollar (SGD) converts alternatively (numerator and denominator) for foreign exchange rate. That is, the basis of SGD/MYR exchange rate was used to analyze in Malaysia while the basis of MYR/SGD exchange rate used to analyze in Singapore. The purpose in using only MYR and SGD for foreign exchange rates is to examine the direct shock from either Malaysia or Singapore. From this method, we can determine and prove that there is bidirectional causal relationship between Malaysia and Singapore foreign exchange rates (Ooi et al., 2009). Malaysia and Singapore are located near to each other and within the Southeast Asia region. Singapore only gained independence from Malaysia in 1965 but it is rated top in terms of international economy developed countries which coming in just behind Hong Kong and United States. Moreover, Singapore shares some partnerships with Malaysia in the political, economic and business sectors. These are the reasons that our study only examines MYR and SGD for foreign exchange rates.

After extracted data from Thomson Reuters Datastream, we transformed both data of exchange rates and stock prices into natural logarithm form. This is because log transformation has taken into account of residuals for the bigger values and it can transform non-uniform residuals become uniform residuals. Nieh and Lee (2001) also supported that the feature of log transformation is useful for analysis of most types of performance and many other measurements. Therefore, the values obtained have been utilized for studying the relationship between stock prices and exchange rates.

3.3 Sampling Design

3.3.1 Target Population in Malaysia

Target Population in Singapore

3.4 Research Instrument

Throughout this research project, statistical analysis software names as EViews 6.0 is employed to analyze the data extracted from Thomson Reuters Datastream. There are various type of tests can be run by EView 6.0 involve t-test , F-test, Augmented Dickey Fuller (ADF) Unit Root test, Johansen Cointegration test and Granger Causality test. Other than that, EView 6.0 also provides the knowledge of whether those variables are in short-run or long-run relationship as well as their causal effect. In general, there are many other software can be used to run various type of tests, but EViews 6.0 is recognized to be more reliable, accurate and familiar to use for us. Moreover, this study also fully utilized two other user friendly software, which include Microsoft Office Excel and Microsoft Office Word.

3.5 Data Processing

3.6 Econometric Techniques Analysis

Multicollinearity

Multicollinearity is a problem that normally exists in time series data or cross sectional data or both. If multicollinearity problem occur in our model, that means we had violate the Classical Regression Linear Model (CRLM). Therefore, our model will not get Best Linear Unbiased Estimator (BLUE). To test for multicollinearity problem, we can use Variance Inflation Factor (VIF). We can calculate VIF as below:

VIF = 1 / 1-R2

If the VIF is higher than 10, that means it is a serious multicollinearity problem. If the VIF is lower than 10, that means it is a low multicollinearity problem. If the VIF is equal to 1, that means there is no multicollinearity problem. To solve for multicollinearity problem, we can increase sample size or develop a new model.

Autocorrelation

Autocorrelation is a problem that normally exists in time series data. If autocorrelation problem occur in our model, that means we had violate the Classical Regression Linear Model (CRLM). Therefore, our model will not get Best Linear Unbiased Estimator (BLUE). To test for autocorrelation problem, there are three types of test can be use. We can use Durbin-Watson Test (Durbin & Watson, 1950), Durbin’s h Test (Durbin, 1970) and Breusch-Godfrey LM Test (Breusch & Godfrey, 1978) to detect for autocorrelation problem. In our research, we will use Bruesch-Godfrey LM Test to detect for autocorrelation problem. We will refer to the F-statistics test and P-value of the F-statistics test. If the P > 0.01, it is means that there is no autocorrelation problem while if P ˂ 0.01, it is means that there is an autocorrelation problem. To solve for autocorrelation problem, we can use Cochrane-Orcutt procedure or increase sample size.

Heteroscedasticity

Heteroscedasticity is a phenomenon normally happens in cross sectional data but it also happens in time series data as well. Heteroscedasticity violates the assumption of the Classical Linear Regression Model (CLRM) that the error term has constant variance.

If there is heteroscedasticity problems occurred, the variance of error term is not achieve at optimal level that it will be overestimated or underestimated. This will lead t-statistic and F-statistic values to be biased in consequences the confidence interval and p-value to be not precise. As the results, the estimated parameters still can be estimated but the estimation result becomes not accurate. This model is considered to be inefficient and not the best model.

There is various types of test that can detect the heteroscedasticity problem, for instances Park Test (Park, 1966), Glejser Test (Glejser, 1969), White Test (White, 1980), Goldfeld-Quandt Test (Goldfeld-Quandt, 1965) and Breusch-Pagan-Godfrey Test (Breusch, Pagan & Godfrey, 1979). In order to detect heteroscedasticity problem in time series data, we use Autoregressive Conditional Heteroscedasticity (ARCH) Test that is developed by Engle (1982).

Johansen co-integration test

Given that all series in the model are co-integrated at same order of Integrated (I), we should conduct the co-integrating relationship between the integrated variables and determine the number of vector co-integrating (r).

For this purpose we employ the Johansen co-integration test. Johansen (1988) as well as Johansen and Juselius (1990) have developed the model as below to examine r.

Where Δ is the difference operator, Y is the variable whose time series properties is examined, α is the constant, β, μ and δ are the coefficients to be estimated, and ε is the white-noise error term.

Considering sensitivity of VAR model to the number of lags of the variables, the optimal lags of the variables of the model must be determined. In this work, we use Akaike Information Criterion (AIC) and Schwarz Criterion (SC) and apply in the VAR model.

Yau and Nieh (2006) and Wu et al. (2012) indicated that there is various methods of estimating co-integration have been used to capture the long-run equilibrium relationship between stock prices and exchange rates. Among these methods, Johansen methodology based on the likelihood ratio with non-standard asymptotic distributions involving integrals of Brownian motions is observed to be the best methodology to proceed with co-integration estimation by Gonzalo (1994).

The elaboration developed by Johansen (1988, 1990, and 1994) are summarized into five VAR models with ECM, which are presented in the following forms:

(1988) (7)

(1990) (8)

(1990) (9)

(1994) (10)

(1994) (11)

To analyze the deterministic term, Johansen decomposed the parameters in the directions of and as, thus we get and . The nested sub-models of the general model of null hypothesis are therefore, defined as below:

Johansen (1994) emphasized the role of the deterministic term, , which consist of constant and linear terms in the Gaussian VAR. Lastly, a decision procedure, follows with Yau and Nieh (2006) as well as Wu et al. (2012) among the hypotheses and of the above five different models, is presented in the following procedure:

We diagnose models one by one until the model that cannot be rejected for the null hypothesis.

Normality test

We would like to find out whether our models satisfy the assumptions of Classical Normal Linear Regression Model (CNLRM). There are several tests of model adequacy to examine whether the model had fulfilled the CNLRM assumptions.

Jarque-Bera test of normality is asymptotically equivalent to the likelihood ratio test (Jarque & Bera, 1980; Jarque & Bera, 1981; Jarque & Bera, 1987). It is implies that it has a large sample, maximum local power. It is also based on the Ordinary Least Square (OLS) residuals. This test first computes the skewness and kurtosis measures of the OLS residuals and uses the test statistic as follow:

where n is sample size, S is the coefficient of skewness, and K is the coefficient of kurtosis. Jarque-Bera showed that large samples. The JB statistic follows the chi-square distribution with 2df under the null hypothesis that the residuals are normally distributed. If the computed p-value of the JB statistic in an application is lower than the significance level, the hypothesis can be rejected and conclude that the residuals are not normally distributed. In contrast, if the p-value is reasonably higher than the significance level, the hypothesis cannot be rejected and conclude that the error terms are normally distributed.

According to Jarque and Bera (1987), the violation of the normality assumption may lead to the use of suboptimal estimators, invalid inferential statements and to inaccurate conclusions. It had highlighted the importance of testing the validity of the assumption.

Model Specification Test

Specification is a process of converting a theory into a regression model. Ramsey (1969) has proposed a general test of specification error known as Regression Specification Error Test (RESET). First, we would estimate the restricted model as Y = β0 + β1X1 + ε and obtain the restricted R2. Next, estimate the unrestricted model as Y = β0 + β1X1 + β2Y2 + β3 Y3 + ε and obtain the unrestricted R2. In the RESET test, F test is used to compute the test statistic value as follow,

If the computed p-value of the RESET test is lower than the significance level, the null hypothesis will be rejecting and can be conclude that the model is incorrect. Otherwise, the null hypothesis will not be rejecting when the computed p-value of the RESET test is higher than the significance level. It can be conclude that the model is correct.

Unit root

The data series we use in this study are time series data. Empirical work based on time series data assumes that the underlying time series is stationary (Gujarati, 2003). But many studies have shown that majority of time series variables are non stationary or integrated of order 1 (Engle and Granger, 1987). Using non stationary time series in a regression analysis may result in spurious regression which was firstly pointed out by Granger and Newbold (1974). Thus before analyzing time series data in an empirical study we should make stationarity test which is commonly done by unit root test. There are a variety of unit root tests used in econometric literature principally Augmented Dickey-Fuller (ADF) test and Phillip-Perron (PP) test. In this study we use ADF unit root test to investigate whether the time series data used in this study are stationary or not. Augmented Dickey-Fuller (1979) test is obtained by the following regression:

Where is the difference operator, are the coefficients to be estimated, Y is the variable whose time series properties are examined and is the white-noise error term.

Granger Causality Test

Once the variables are established which are non-stationary and co-integrated, Vector Error Correction Model (VECM) is the adequate method to examine the causal relations between the exchange rate and stock price index (Granger, 1988). When the variables are co-integrated, then there exists an error-correction representation of the form as below:

where ∆EXt is denotes of changes in the exchange rate and ∆SPt is changes in the stock price. ECTt-1, which is SPt-1 – γEt-1, is an error correction term derived from the long run cointegrating relationship. The error correction term can be predictable by using the residual from a cointegration regression. The estimated δ1 and δ2 explained the speed of adjustment.

In the absence of any cointegrating relationship between the variables, a Vector Autoregression model is used (Granger, 1969). The standard granger causality test to examine whether there exist feedback, bi-directional or unidirectional causality between variables. The Granger method (Granger, 1988) seeks to find out how much of a variable, Y can be explained by past results of Y and whether adding lagged value of another variable, X can improve the justification. To test the causal relations between exchange rate and stock price can be based on the following bivariate autoregression:

where EX is exchange rate variables, SP is stock price variables and ε is assumed to be serially uncorrelated with zero mean and finite covariance matrix. Based on Granger (1969), imply that for EX to Granger-cause SP, the coefficient βk ≠ 0 in Equation ( ), whereas λk = 0 in Equation ( ); and for SP to Granger-cause EX, λk ≠ 0 while βk = 0. If we allow for the possibility of j = 0 in the addition symbol for Equations ( ) and ( ), the relationship between the two time series is said to be instantaneous (Abdalla & Murinde, 2010).

To decide whether or not to reject the Granger causality test, Granger applies the F-test of overall significance, and had computed as:

Where RSSUR is the unrestricted residuals sum of squares from the estimated Equation ( ) [or Equation ( )] and RSSR is the restricted residuals sum of squares from the estimated Equation ( ) [or Equation ( )] under the null hypothesis. F test follows F distribution with m and (n – k) df. In the present case m is equal to the number of lagged M terms and k is the number of parameters estimated in the unrestricted regression. If the computed F value exceeds the critical F value at the chosen level of significance, we reject the null hypothesis, in which case lagged M terms belong in the regression. This implies evidence of causality. On the other hand, the null hypothesis is not rejected implies causal relationship.

There are four possible results of this Granger causality test which are unidirectional causality from SP to EX; unidirectional causality from EX to SP; feedback causality between EX and SP, and independence between EX and SP (Abdalla & Murinde, 2010).

Conclusion

In conclusion, this chapter discusses on how the whole research was carried out, it provides a briefly description on the plan of the research was done to complete the work. In this research, Eviews was used to perform the particular tests in order to detect the particular problems and the actual results of the analysis will be conducted in the following chapter.