Excerpt

## TABLE OF CONTENTS

**CHAPTER ONE-INTRODUCTION **

1.0 BACKGROUND AND INTRODUCTION

1.1 AIMS AND OBJECTIVES

1.2 SCOPE AND METHODOLOGY

1.3 FINDINGS

**CHAPTER TWO- LITERATURE REVIEW **

2.0 INTRODUCTION

2.1 MACROECONOMIC VARIABLES AND STOCK PRICES

2.2 MACROECONOMIC VARIABLES AND STOCK PRICES IN DEVELOPING MARKETS

2.3 MACROECONOMIC VARIABLES AND STOCK PRICES IN DEVELOPED MARKETS

2.4 HEDGING STRATEGIES FOR MOVEMENTS IN MACROECONOMIC VARIABLES

**CHAPTER THREE-DATA AND METHODOLOGY **

3.0 INTRODUCTION

3.1 STRUCTURE

3.2 DATA

3.3 MAIN HYPOTHESIS

3.4 STATISTICAL TESTS

3.5 LIMITATIONS

**CHAPTER FOUR-EMPIRICAL RESULTS AND ANALYSIS **

4.0 INTRODUCTION

4.1 IMPACT OF CHANGES IN MACROECONOMICS IN DEVELOPED MARKET

4.11 NORMALITY TESTS

4.12 CORRELATION

4.13 UNIT ROOT TEST

4.14 GRANGER CASUALTY

4.15 JOHANSEN CO-INTEGRATION TEST

4.16 VECTOR ERROR CORRECTION MODEL

4.17 WALD TEST STATISTIC-TESTING SHORT RUN

4.18 MULTIPLE REGRESSION MODEL

4.19 IMPACT OF CHANGES IN MACROECONOMIC VARIABLES IN DEVELOPING MARKETS

4.20 NORMALITY TEST

4.21 CORRELATION

4.22 UNIT ROOT

4.23 GRANGER CASUALTY TEST

4.24 JOHANSEN CO-INTEGRATION TEST

4.25 VECTOR ERROR CORRECTION MODEL

4.26 WALD TEST FOR SHOT RUN EFFECT

4.27 MULTIPLE REGRESSION MODEL

**CHAPTER 5-HEDGING STRATEGY IN DEVELOPING AND DEVELOPED MARKETS**

5.0 INTRODUCTION

5.1 DIVERSIFICATION WITH A RISK FREE ASSET

5.2 DIVERSIFICATION USING GOLD FUTURES

5.3 DIVERSIFICATION WITH AGRICULTURE FUTURES (LONDON WHEAT FUTURES)

**CHAPTER 6-CONCLUSION **

6.0 CONCLUSION

6.1 AIM OF THE STUDY AND OBJECTIVES

6.2 FINDINGS

6.3 RECOMMENDATIONS

**CHAPTER 7: REFERENCES **

**CHAPTER 8-APPRENDIX **

APPENDIX A -DEVELOPED MARKET STATISTICAL TESTS

APPENDIX B: DEVELOPING MARKET STATISTICAL TESTS

## ABSTRACT

The purpose of this study is to analyse the changes in macroeconomic variables and evaluate the impact on a company’s stock prices, by examining the impact of changes macroeconomic variables, determining which macro-economic variables that have the least and most impact on stock prices and also suggest ways in which the impact on the macroeconomic variables on stock prices can be hedged against using agricultural futures, metal futures or a risk-free asset.

The study will use five econometric models to test this impact, these include the Granger Causality test, Johansen Co-Integration test, Vector Error Model, Walt Test statistic, Multiple Regression Model. A review of a number of academic literature by notable analysis for both developed and developing markets will be provided. The FTSE share price index will be used in the study to represent the developed markets and the JSE share price index will be used in the study to represent the developing markets.

The results provided are mixed with the Johansen Integration and Multiple regression tests indicating presence of this impact on stock prices by macroeconomic variables. The Vector Error correction Model, Wald Test statistic and Granger Causality providing the opposite results. Based on the multiple regression model, interest rates having a negative impact, inflation, currency exchange and unemployment rate having a positive impact on stock prices in developed markets whereas in the developing markets the currency exchange rate and the interest rate had a negative impact on stock prices where as inflation had a positive impact. The developing and developed market’s interest rates had the greatest impact and currency exchange rate had the least impact on stock market indices.

## ACKNOWLEDGEMENTS

I would want to express my most profound gratitude to all individuals who assisted me in fulfilling this study, an uncommon appreciation to my supervisor Dr Kudakwashe Bondamakara and Dr Muhammad Akbar

## CHAPTER ONE-INTRODUCTION

## 1.0 BACKGROUND AND INTRODUCTION

The UK has experienced a lot of instability in different macroeconomic variables since the result of the referendum to exit the European Union was announced on the 24th of June 2016. The Brexit decision also led to high volatility in stock prices (Ramiah, Pham and Moosa, 2017). Stock markets globally have of late been more connected than ever before this has often led to the effects of any changes being felt right across all the stock markets (Forbes and Rigobon, 2002).

According, to Rodionova, Cox and Chu (2017) the Pound Sterling depreciated against the United States Dollar by about 14% as a result of the decision to exit the EU. Also to be affected by this result was the UK FDIs which fell by about 0.9%, inflation for the country rose from 0.3% to 1.2% during the last quarter of 2016 and the official bank rate plunged by 50% to 0.25%.

A number of notable analysts have provided literature on stock price movements and how their determined. The main proponent of stock price movement/Efficient Market Hypothesis was Fama (1970) who suggested that any stock prices will change on the account of news which is unpredictable and any expected news/information is always included in the asset prices. Sharpe (1964), Lintner (1965) and Mossin (1966) invented one of the first viable asset pricing theories which used stock market index to explain an investors returns, known as CAPM (Capital Asset Pricing Model). CAPM uses just one factor, namely stock market index, in order to explain common stock returns. The CAPM model was heavily criticised because of its inability to include other factors to determine investors return. Ross (1976) introduced a multi-factor model and then introduced a multi-factor model known as Arbitrage Pricing Theory which included a few macroeconomic variables in his study.

Elton and Gruhen (1991), suggested that if expected returns affect share prices, hence share prices are determined by macroeconomic variables, because expected returns rely heavily on macroeconomic variables. Tobin (1965), suggested that inflation and real returns on Treasury bills (T-bills) were negatively correlated with stock prices. French et al (1987), concluded that interest rates had no influence on stock prices. Sharma and Mahendru (2016) went on further to say that not only do macroeconomic variables influence share prices but also the countries’ policies. McQueen and Roley (1993), suggested that if macroeconomic variables were favourable, stock market indices performed negatively.

### 1.1 AIMS AND OBJECTIVES

The aim of this research is to analyse the changes in macroeconomic variables and evaluate the impact on a company’s stock prices. The objectives of this study are:

1. To examine the impact of changes in macroeconomic variables (Interest rates, inflation, gross domestic product and unemployment rates) on stock prices;

2. To determine which macro-economic variables has the least and most impact on stock prices

3. To suggest ways in which the impact on the macroeconomic variables on stock prices can be hedged against.

In order to help achieve the objectives stated above, the following research questions have been formulated:

1. What is the impact of changes on macroeconomic variables on stock prices?

2. Which macro-economic factors have the least impact on the stock prices?

3. Which macro-economic factors have the greatest impact on the stock prices?

4. Which hedging strategies can be adopted to mitigate the impact of macroeconomic variables on stock prices?

### 1.2 SCOPE AND METHODOLOGY

This study will focus on FTSE Indices (UK) and JSE Indices (South African). The macroeconomic indicators to be used will include, Currency, inflation rate, Interest Rates and Unemployment rate.

The scope of the study will be as follows firstly, the study will review academic literature to show gaps in my research, to demonstrate an understanding of the research on macroeconomic variables, stock prices and hedging strategies, generate a new null and alternative research hypothesis on whether there is a relationship between the macroeconomic variables and stock prices and summarise and evaluate past research indicating any similarities/differences. Secondly, the study will move onto discuss the methodology used to obtain data and information about the study and also carry out statistical the tests on extracted data. Lastly, the study will evaluate/analyse the findings from the tests carried out, recommend any solutions/hedging strategies and provide a conclusion.

### 1.3 FINDINGS

The findings indicated mixed results based on the models used in the study. The granger causality tests indicated that out of all macroeconomic variables in the UK the unemployment time series data could be used to predict the FTSE, in the developing markets interest rates and the JSE share price index could predict each other. The co-integration test in both developing and developed markets indicated that there was some form of co-integration between the stock price indices and the macroeconomic variables in their respective countries. The vector error correction model indicated that there was no long run relationship in both markets between the stock price indices and the macroeconomic variables. The Wald test statistic was there was no short run impact on the stock prices by the macroeconomic variables in either markets. The multiple regression model contradicted all models and reflected that in fact all changes in macroeconomic variables had an impact on stock prices, with interest rates having a negative impact, inflation, currency exchange and unemployment rate having a positive impact on stock prices, In the developing markets the currency exchange rate and the interest rate had a negative impact on stock prices where as inflation had a positive impact. The developing and developed market’s interest rates had the greatest impact and currency exchange rate had the least impact on stock market indices.

## CHAPTER TWO- LITERATURE REVIEW

## 2.0 INTRODUCTION

This section will provide detailed review of previous academic literature to demonstrate an understanding of the topic. A general overview of the impact of macroeconomic variables and stock prices will be provided, followed by the relationship in developing and developed markets.

### 2.1 MACROECONOMIC VARIABLES AND STOCK PRICES

A number of pioneers have studied the relationship between macroeconomic variables and stock prices. Most of the studies have were from the 20 **th** century with the 21 **st** century lacking much research on the topic, hence the reason for this study.

Macroeconomics is a division of economics which assesses the impact of choices on the major aggregate level of the economy, where aggregates include the consumer price index(CPI)/Inflation/deflation, money supply(M1)/(M2), currency movements, gross domestic product, interest rates and unemployment level (Marshall, 1895). On the other hand, shares signal a sense of ownership of stock in a company, and stock prices/share prices is an indicator of the success of the business (Van Horne and Wachowicz, 2008).

According to Elton and Gruber (1991), share prices are determined by the expected return and expected cash flows. Hence macroeconomic variables would influence share prices as expected returns rely heavily on changes in other macroeconomic variables as suggested by Chen, Roll and Ross (1986) study.

According to Tobin (1965) suggested that inflation and real returns on treasury bills were negatively correlated with stock prices. Tobin’s (1965) study was supported by Fama (1991). A further study by Geske and Roll (1983) confirmed this relationship although argued that there was a positive relationship between the stock price and economic activity.

Research has indicated that short term and long term interest rates do not have an impact on stock returns French et al, (1987).

Chen, Roll and Ross (1986) pioneers of the APT concluded that macroeconomic variables had an impact on asset prices/returns these included differences between the spread of long and short term interest rate; production growth in the industry, difference between the inflation.

Sharma and Mahendru (2010) suggested that not only do macroeconomic variables affects stock prices but also, changes in other countries policies could have an impact as stock market integrates. It can be argued that share price fluctuations are not only or limited to changes in macroeconomic variables but also other factors and hence this would suggest that more variables should be included in any research to produce sufficient evidence.

McQueen and Roley (1993) further on went to suggest that if the macroeconomic factors are favourable (i.e.) favourable interest rates for borrowers, investors’ stock markets indices perform negatively. This implies that stock prices are likely to plunge due to investors hunting for favourable interest rates offered by banks and would reduce the liquidity of stock markets.

Apergis and Miller (2009) study concluded that the effect on global demand, and oil prices had insignificant effect on stock prices/returns which could suggest that other macroeconomic variables such as interest rates & exchange could have an significant impact on stock market prices/returns. This implies that the stock markets are not yet entirely integrated, also caution should be taken as no test were performed to compare their hypothesis or prove it.

### 2.2 MACROECONOMIC VARIABLES AND STOCK PRICES IN DEVELOPING MARKETS

Ibrahim and Aziz (2003) studied the relationship between stock prices and macroeconomic indicators in Malaysia; they indicated that changes in exchange rates had a negative impact on stock price whereas changes in money supply had a positive effect in the short term where as in the long term there was a negative impact on the stock prices. In India exchange rates and gold prices have a high impact on stock prices (Sharma and Mahendru, 2010).

Research by Kandir (2008) on Turkish portfolio returns argued that stocks are definitely affected by inflation rates, exchange rates and interest rates. On the other hand, stock returns are hardly affected by the domestic production, oil prices and the money supply. Although the study is useful it would be naive to conclude that this research would be accurate globally as more markets are integrated, and hence why more examining more markets would have given an accurate picture for global investors. Kandir research also indicates that the variables which are correlated could affect the stock prices/returns in a similar way.

Enisan and Olufisayo, (2009) examined the causal relationship between the economy and the stock markets of seven sub-saharan Africa, it indicated that the two variables were co-integrated and that the stock market caused economic growth mainly in Egypt and South Africa, contrary in Cote D’Ivoire, Zimbabwe, Kenya and Morocco. This suggests that if the South African economy were underperforming this would be reflected in the stock prices whereas in the Zimbabwean economy it would be the opposite.

Kuwornu (2011) study of macroeconomic variables and stock returns in Ghana indicated CPI had a positive impact on stock price returns whereas exchange rates and Treasury bill rate negatively affected stock market returns. Adam and Tweneboah (2008) also performed a further research on Ghana and discovered that there was long run relationship between the macroeconomic variables and stock, with specific reference to interest rates and inflation with exchange rates having a short run relationship. This also supports Kuwornu’s (2011) evidence that the two variables were co-integrated.

Evidence from Latin America (Brazil, Chile, Argentina and Mexico) indicated that the foreign exchange market (exchange rates) was related to the stock prices which could have been as a result of the linkage with the US Stock Market (Diamandis and Drakos, 2011).

Evidence from Pakistan between its macroeconomic variables and the KSI index indicated that there was co-integration with inflation having the largest negative impact on the stock prices (Nishat, Shaheen and Hijaz, 2005). Hence, any increase in inflation would result in the plunging of stock prices and vice-versa. This evidence can also be supported by Pramod Kumar, Naik and Puja, Padhi (2012) study that the Indian stock market index and five macroeconomic variables(Money supply, treasury bill rates, exchange rates, industrial production index and wholesale price index) are co-integrated, with money supply positively affecting the stock prices and inflation impacting stock prices negatively.

Basher and Sadorsky (2006) research on emerging markets indicated that the oil price risk factor had a large impact on stock returns. This suggests that not only the common variables such as exchange rates or interest rates have an impact on stock prices but also other factors.

In Lithuania money supply and gross domestic product(GDP) have a positive impact on stock price on the other hand short term interest rates, exchange rates and unemployment rate have a negative impact (Pilinkus and Boguslauskas, 2009).

Maysami, Howe and Rahmat (2005) performed a study on Singapore stock market and macroeconomic variables which indicated co-integration of macroeconomic variable movements with stock prices. Wongbangpo and Sharma, (2002) performed a further study on the interaction between macroeconomic variables and stock prices in five ASEAN countries (Thailand, Singapore, Indonesia, Philippines and Malaysia) and they concluded that they co-integrate but also recommended that viable government policies on the economic could improve the returns in the macro economy and stock price markets.

Hussainey and Khanh Ngoc (2009) study on Vietnam went on further to support evidence of co-integration of stock price movements in relation to the macroeconomic variables, but also concluded that the US macroeconomic variables had a great impact on Vietnam’s stock prices

### 2.3 MACROECONOMIC VARIABLES AND STOCK PRICES IN DEVELOPED MARKETS

Granger, Huangb and Yang (2000) performed research on the United Stock market and discovered that changes in the exchange rate have little to no impact to asset prices/returns. Hence this suggests that more tests have to be performed to ensure that this relationship is actually appropriate between the two variables. The unsatisfactory aspect of their study was only centred in the United States and hence can be only concluded for the United States stock Market.

Granger, Huangb and Yang (2000) also went on to suggest that since capital markets became increasingly integrated changes in either the stock prices or exchange rates could be reflected more in the current account balance due to movement in capital. Hence in the research more caution has to be taken as a change in one macroeconomic variable would also affect another and not necessarily stock prices. Thus, if there are lower interest rates, it could also suggest that the domestic currency is less demanded by international investor and hence this would suggest that the local currency will depreciate due to lack of capital inflows (Granger, Huangb and Yang, 2000).

In the USA, domestic production, yield curve and risk premium were found to have a significant impact on stocks (Chen, Roll and Ross, 1986). In the UK, changes in retail price index and the lending rates indicated great impact on the stock prices/returns (Clare and Thomas 1994). In Singapore, research suggests that money supply do have an impact on stock prices (Mookerjee and Yu, 1997). In Japan, inflation rate, lending rates, money supply and exchange rates had a significant impact on stock prices (Mukherjee and Naka, 1995).

Despite more stock markets being integrated less research has been gathered on the effect of changes in macroeconomic variables in one country and its impact of stock prices/returns in another country, thus most research has been concentrated in the resident’s country. Nieh and Lee (2001) tried to perform a research on the G-7 countries of that kind, they discovered that depreciation of a currency in Germany caused a downward trend on its stock prices/returns; on the other hand in the UK it tends to actually do the opposite. This indicates negative correlation of stock price movement or performance. However, Kurihara’s (2006) research suggested that there was no relationship between stock prices and macroeconomic variables, although there was interdependence between the US and Japan Stock prices that are any stock price movements in the US would impact those in Japan.

In New Zealand the NZSE40 indicated that it was in fact determined by the gdp, money supply and interest rates, but the NZE40 did not play a role in making any changes in the macroeconomic factors during the period of 1990 to 2003 (Gan et al., 2006).

Kurihara Yutaka (2006) investigation of the relationship between Japan’s macroeconomic variables and stock prices indicated that interest rates had no influence on the stock prices on the other hand exchange rates did have an impact. This assertion was further supported by research on ten European countries macroeconomic variables (inflation, interest rates, imports and employment) in relation to stock price which indicated an inverse relationship with results of the association being significant in the United Kingdom (Asprem, 1989).

Evidence from the USA between 1974 and 1998 indicated that the gross domestic product and the S&P 500 were positively related with inflation, exchange rates and interest rates having the opposite relationship (Kim, 2003).

Maysami and Koh (2000) performed a study on Singaporean stock prices and selected macroeconomic variables and they discovered the existence of co-integration, with exchange rates and interest rates being the most sensitive.

Alam and Uddin, (2009) performed research on stock prices and interest rates for fifteen developing and developed economies some which include Bangladesh and Germany, and they concluded that interest rates have a negative relationship with the share prices. This could suggest that relationships between macroeconomic variables and stock prices are similar in developing and developed markets.

An unusual study of the United Kingdom indicated a positive relationship between crude oil prices and the stock price values, with the volatility on the oil prices having a direct impact on the share values of the sector (El-Sharif et al., 2005).

### 2.4 HEDGING STRATEGIES FOR MOVEMENTS IN MACROECONOMIC VARIABLES

With so much risk in the financial markets, in the 20 **th** century a few pioneers invented models to reduce risk. Firstly, Harry Markowitz proposed a mean-variance model, which later on led to a two asset hypothesis which tried to diversify our risk by distributing different assets in a portfolio, in addition to that he also proposed an efficient portfolio model which identified the most optimal combination of assets in a portfolio to provide minimum risk for high returns (Dobbins, Witt, and Fielding, 1993). From then on many models further built onto his work such as the CAPM and Sharpe ratio which tried to minimise the risk. James Tobin suggested the adding a risk free asset to reduce the risk. It should be noted that all these models would not play a vital role in today fast paced modern world with technology changing every day. Although we cannot entirely neglect their work but to build on what has been provided.

According to Zanotti, Gabbi and Geranio (2010) research resulted in reduction of variance through using futures hedging in electricity portfolios. As futures are quite risky as you are locked in a contract and less in control of the spot prices and hence it would have been ideal if they tested not only electricity portfolio’s but other industries for example food industry. Poomimars, Cadle and Theobald (2003), took an extra mile by assessing 7 markets and their results suggested that dynamic-futures hedging rations led to improvement in hedging performance and hence this supports Zanotti, Gabbi and Geranio (2010) which was done in only one market.

Galati and Melvin (2004), offered a strategy for hedging interest rates , that is by investing in currencies with high returns through borrowing low interest rate currency for example the US$ and taking a buy and hold position in a currency with a higher interest rate such as the Australians in order to offset any changes in interest rates currencies.

Morey and Simpson (2001) carried out an investigation on Canada, Germany, Japan, Switzerland and United Kingdom between1989-1998 and discovered that they were two strategies that were either to hedge when the forward rate was at a premium and if it wasn’t not to hedge. Hence this suggests that one can hedge using the purchasing power parity model to hedge against risk.

Lacoviello and Ortalo-Magné (2003) provided a solution to reducing risk; their suggestion was that investing in the property market in London provided favourable returns and risk instead of investing in financial assets. Although property investment is safer to derivative investments, 5 years later after their research one of the worst financial crisis 2008 caused by housing bubble occurred and which made companies and households bankrupt and hence investments such as Bonds would have been considered safer to real estate.

According to Garber and Spencer (1995) interest rate hedging also depends on the timing and size of the funds being invested thus opposing those researchers and analysts who assume that derivatives do not carry any costs. During this review most researchers hardly included such forms of limitations, which would impair the accuracy of their results.

Bessembinder (1992) examined the pricing of risk for assets and futures and discovered returns on agriculture futures and foreign currency vary with the net holdings of hedgers when the systematic risk is controlled. He went on to conclude that the hedging pressure was a determinant on the premium paid on futures. This implies that the cost of obtaining futures is never fixed but varies with regards to demand in that market segment and hence cheaper cost of obtaining futures will be realised in the market where there is little hedging pressure.

Massa and Simonov (2006) opposed the idea of hedging and their research indicated that investors are much more likely obtain higher returns through investing in stocks that they are familiar with (Familiarity based investment) than by hedging. Also Hedging for the purposes of increasing firm value, cash flow and sales revenue could put financial strain on a firm because most hedging strategies involve borrowing therefore not hedging will be the ideal decision of a financial manager (Mello and Parsons, 2000).

Broll and Wong (2010) concluded if any bank is to have an effective hedging strategy they would have to take into consideration the business cycle because their risk behaviour was related to it. This suggests that one is bound to likely to be risk averse during a recession and hence hedging little to nothing. During a boom or growth period one is likely to be risk loving and thus investing in high risk investment such as futures.

One disadvantage of hedging is that it is used for speculative purposes and not usually for hedging purposes by financial traders (Wei, 1999). This could suggest that more risk is added in trying to defend volatility from stock prices.

Depending on future contracts to hedge against exchange rate fluctuations can be seen as a defence against adverse effects and a better approach to saving investments if the worst comes to the worst, that is recessions (Röthig, Semmler and Flaschel, 2007). This implies that despite the risky side of hedging instruments, if the overall benefit of risk management by corporate firms could become fruitful this would actually sustain economies at the macroeconomic level and not just at corporate level.

Adrangi, Chatrath and Raffie (2003) argued that inflation had a no impact on gold price and silver although expected inflation was likely to have an impact on the prices of gold. This implies that silver and gold would be one of the best hedges for inflation either in the long run or short run.

Research has indicated that the hedging of a company’s cash flow increase the firm’s value on the other hand this would increase the chances of the firm increasing their reliance on banking funds. It can be argued that although investors would benefits from an increase in their shareholder wealth, the company’s financial risk would increase, hence more pre-cautionary the firm would have just avoided the macroeconomic risks but increased their own specific risk (Disatnik, Duchin and Schmidt, 2013).

Soenen and Lindvall (1992) suggested that if US investors could hedge their risk through diversifying internationally and further reducing their risk through currency hedging, although it would cause a reduction in returns for the investors. This is a good initiative as the firm or investor would avoid unnecessary risk although the depletion in returns could be unsatisfactory, hence opting for other low risk investments such as bonds would be ideal.

Research by Yaganti and Kamaiah (2012) suggested that the efficiency of hedging with agricultural futures was less than that of base metals. Base Metals have fewer suppliers than agricultural products which mean less volatility and hence it would be ideal to hedge using base metals.

## CHAPTER THREE-DATA AND METHODOLOGY

## 3.0 INTRODUCTION

The research will consist of mainly of quantitative analysis to draw a conclusion on whether macroeconomic variables have an impact on stock prices. This section will provide research methodology and the data to be used in the study. Also the structure of the dissertation will be provided.

### 3.1 STRUCTURE

The study will focus on reviewing any academic literature in the past and provide a better understanding of the relationship between macroeconomic variables and stock prices and hedging strategies that have been adopted in the past. The literature review will also include some theory that has been included in the past to elaborate on other aspects. The academic literature will not be only limited to the stock markets that are being assessed for this study.

A range of statistical tests will be performed on both variables and an analysis will be provided for each outcome. These will include Unit Root tests, Co-integration tests, Granger Casualty test, Multiple Regression Model, ARMA and the GARCH test. These will be discussed in more contexts below. The first four tests will be performed to determine whether the macroeconomic variables have an impact on the stock prices and vice-versa, the last two statistical tests will determine how best an investor can hedge the movement in stock prices caused by changes in the macroeconomic variable.

The empirical results will be discussed in more detail and observing whether there is any links to the relationships with previous academic literature provided in the first part of the stud.

Lastly, a conclusion will be provided on the statistical tests and its analysis stating whether the macroeconomic variables have an impact on the stock price of either the FTSE Index or the JSE Index, also stating which variables has the most and least impact on stock prices. A recommendation will be provided for how to hedge the fluctuation of stock prices based on the changes in macroeconomic variables. A list of references and appendix will be provided at the end of the study to further enhance the information.

### 3.2 DATA

The study will obtain data for quantitative analysis published by the Bank of England, ONS, Johannesburg Stock Exchange, London Stock Exchange and Investing.com. The data sample period will range from 2010-2017 analysing approximately 86 monthly observations of the FTSE and JSE market index and their respective macroeconomic variables.

Abbildung in dieser Leseprobe nicht enthalten

Fig.1 OBSERVATIONS

The data obtained above is equivalent to 86 monthly observations for each variable. The FTSE Share price denominated in £ with the JSE Share price in Rands. For the other Variables such as inflation, exchange rate and interest rates will be presented as a percentage.

The FTSE Share price index will be presented as a weighted average stock price of all the companies listed on the London Stock Exchange. The FTSE share price index has also been chosen as a developed market indices as I believe that this provides us with a reflection of other developed markets such as the Frankfurt Stock exchanges and many others. It is also widely diversified amongst many industries such as the transport, finance and manufacturing industries.

The JSE Share price index will be presented as weighted average stock prices of all the companies listed on the Johannesburg Stock Exchange. The JSE Share price index will be chosen as the developing market representation as this fully reflects performance of other developing markets. It is diversified amongst many sectors just like the FTSE and has a linkage with the FTSE as many of its companies are listed there.

The inflation, unemployment rate and interest rates will be presented as percentages. These variables will be used in the study/statistical tests to identify whether a relationship exists between the macroeconomic factors and its share prices. Each respective currency’s performance will be compared to the US$ as this is one of the most global currencies.

The sample period as seen in Fig 1 for all variables and stock price indices will range from 1/07/2010-30/06/2017 because of the dramatic changes that have occurred over years such as BREXIT and the changes in Official bank rate by the Bank of England.

Abbildung in dieser Leseprobe nicht enthalten

Fig 2. FTSE SHARE PRICE AND JSE SHARE PRICES

Both the FTSE and JSE share price indices have experienced some upwards and downward trends over the past seven years. As seen from Figure 2, the movements in the share prices are quite identical to each other. The FTSE and JSE share price reached an all-time high in 2017, whereas the FTSE reached its lowest in 2011 and the JSE in 2010. This could have been as a result of the ripple effects of the economic recession in 2008.

Abbildung in dieser Leseprobe nicht enthalten

**Figure 3.** FTSE AND JSE SHARE PRICE RETURNS

The FTSE and JSE returns are highly volatile with the JSE returns having a peaking of 8% in 2012 and 2013 and FTSE returns having a peak of approximately 5.5% in 2011. The lowest performance of FTSE ranged between -5 to -6% in 2011 and 2012 with JSE returns hitting a low of -5% in 2013.

Abbildung in dieser Leseprobe nicht enthalten

**Fig 4** Macroeconomic variables for the United Kingdom and South Africa.

The macroeconomic variables of both the United Kingdom and South Africa have not been as volatile as their respective share price index. As seen from figure 4 UK’s Inflation has been increasing rapidly whilst South Africa’s inflation has been decreasing over the 2 years, Both the GBP and ZAR currencies have depreciated against the dollar with the Interest plunging in the last few years. This data will be used to identify whether the changes of these macroeconomic variables have had any impact on stock prices over the last seven years.

### 3.3 MAIN HYPOTHESIS

The main hypothesis for this study will be as follows;

Abbildung in dieser Leseprobe nicht enthalten

### 3.4 STATISTICAL TESTS

a) NORMALITY TEST

The first tests would involve testing whether the time series data we are provided follows a normal distribution; hence the jarque-bera measurement and kurtosis will give us an idea about this. The time series data for the macroeconomic variables will be combined, whereas the stock prices data will be done individually.

**b) CORRELATION (MULTICOLLINEARITY)**

A multi-correlation test will be performed to understand the relationship between the macroeconomic variables themselves before testing the impact of their impact on the stock prices. This will give us a true picture when concluding our findings.

It should be noted that positive correlation is equivalent to 1, no correlation (0) and negative correlation -1.

[Formula]

c) UNIT-ROOT TEST

In the research a unit root test will be used to identify whether the time series data for the variables has possesses a unit root or not. It should be noted that if the time series possess a unit root it would reflect that the data is non-stationary or the market’s data being tested is not efficient and vice-versa. The augmented dickey full test will be performed to give us the results on whether the time series is stationary or non-stationary, through a t-static.

Abbildung in dieser Leseprobe nicht enthalten

The null hypothesis for this statistical test would be a unit root and the alternative hypothesis is a non-unit root.

Abbildung in dieser Leseprobe nicht enthalten

d) GRANGER CASUALTY TEST

Following the unit root tests, a granger casualty to co-integration test/_multi-variant maximum likelihood test will be performed to identify whether the macroeconomics data and the stock price data are co-integrated/ whether there is a long run equilibrium/not between the macroeconomic variable and the stock prices.

[Formula]

Since this statistical test can only be performed on two variables, each variable will be tested against the stock prices, take for example:

CPI VS FTSE INDEX PRICES OR GDP RATE VS JSE INDEX PRICES

The null hypothesis for the statistical test is that the data sets are co-integrated and the alternative hypothesis is that the two data sets are not co-integrated.

Abbildung in dieser Leseprobe nicht enthalten

Hence if the data is co-integrated then we can conclude that either of the two data sets does not affect each other

e) MULTIPLE REGRESSION MODEL

This model will give an overall outlook on whether the macroeconomic variables in actual fact impact the stock returns of the selected stock indices in either the developed and developing markets. This test is similar to that of the Arbitrage Pricing Theory. This will also give us an idea on which macroeconomic variable has the least and greatest impact on the stock prices/returns

[Formula]

The hypothesis null hypothesis would be yes they impact the stocks and the alternative hypothesis will be no the variables do not affect

Abbildung in dieser Leseprobe nicht enthalten

FORECASTING/HEDGING

After performing the statistical test that determine the relationship between the macroeconomic variable and stock price, an autoregressive moving average (ARMA) and General Autoregressive Conditional Heteroskedasticity test will be performed to identify whether we can forecast the data and hedge effectively.

1. ARMA (0,0) to (5,5)

An ARMA tests will be performed to identify the suitable time series process and hence it can be used to predict the future values in the series. By performing this test, we can suggest to the investor whether dynamic hedging or static hedging will be suitable in either index. First part will involve provide autocorrelation and partial correlation function which would give us an idea on the appropriate forecasting models to use. [Formula]

Secondly, we will tests at most 36 models (0, 0) to (5, 5) using the Akaike and Schwarz information criteria to find the best model for forecasting. Lastly, we will forecast using the suitable model and provide whether dynamic/static hedging is suitable or not, the spot and future prices of a range derivative or commodity will be tested.

Abbildung in dieser Leseprobe nicht enthalten

2. ARCH and GARCH TEST

Lastly, Autoregressive Conditional Heteroskedasticity (ARCH) and General Autoregressive Conditional Heteroskedasticity (GARCH) tests will be performed to analyse the volatility of the stock prices and impact of future volatility caused by any changes of the macroeconomic variable.

The Garch tests will be divided into GARCH (1, 1), T-GARCH, EGARCH and GARCH-M which will be further explained in the analysis. These tests will use the t-statistic/p-value to evaluate the findings hence either rejecting or accepting the alternative or null hypothesis.

Abbildung in dieser Leseprobe nicht enthalten

Where the variance equation is:

Abbildung in dieser Leseprobe nicht enthalten

### 3.5 LIMITATIONS

A number of limitations could arise whilst performing this statistical test that is human error/mistakes and problems that arise from the digital age of using technology and hence one should take careful considerations when reading into the analysis.

## CHAPTER FOUR-EMPIRICAL RESULTS AND ANALYSIS

## 4.0 INTRODUCTION

The empirical results and analysis will be divided into two section, developed and developing market analysis. At the end of the analysis of both markets a hedging strategy will be provided for both developed and developing markets.

### 4.1 IMPACT OF CHANGES IN MACROECONOMICS IN DEVELOPED MARKET

#### 4.1.1 NORMALITY TESTS

Abbildung in dieser Leseprobe nicht enthalten

STANDARD DEVIATION

Standard deviation is a measurement tool used in statistics to measure the variance in returns of variable (Brooks, 2015). The standard deviation is rather stable for official bank rate and the GBP/USD at 8% and 12% respectively which is an indicator that there is not much movement or change in these variables. On the other hand, the inflation and unemployment rate have deviation of approximately 122% and 136% respectively, this implies that they have frequently changed over the past 10 years and could have created high uncertainties in the economic environment.

SKEWNESS

Skewness is an asymmetry measurement tool of the probability distribution of a random variable’s mean (Brooks, 2015) [Paraphrase]. All four variables have a negative Skewness which is an indicator that the median is greater than the mean. Hence the average value of each of these variables is less than the median, which could indicate decline in each variable.

KURTOSIS

Kurtosis measures the peak returns in a distribution (Verbeek, 2015) [Paraphrase]. All the kurtosis values are positive which is an indicator that the data obtained on the four variables will not be extreme if it changes at all.

JARQUE-BERA

The Jarque Bera probability for all variables except inflation are below zero/zero indicate that the changes in their values are not normally distributed and hint a sign of inefficiency, hence this will mean that first difference test of the unit root test would have to be carried out to make the data stationary.

#### 4.1.2 CORRELATION

Abbildung in dieser Leseprobe nicht enthalten

The inflation has had a negative correlation with other variables which is an indicator that there is a negative relationship between it and the variables. This is ideal as the result is unlikely to be the same.

The GBP/USD currency only has a negative correlation with the inflation; on the other hand it has an almost positive correlation with the official bank rate and unemployment rate which is an indicator that they have a positive relationship. Hence this could suggest that if currency is to have an impact on stock price and so will the official bank rate and unemployment rate.

Official bank rate only has a negative relationship with the CPI where as a positive correlation with the GBP/USD and the unemployment rate.

The unemployment rate has a positive relationship with all variables except the inflation.

**[...]**

- Quote paper
- Kudzanai Chakona (Author), 2017, Changes in Macroeconomic Variables and Their Impact on Stock Price Indices. A Case Study of the Financial Times Stock Exchange (FTSE) and Johannesburg Stock Exchange (JSE) Indices, Munich, GRIN Verlag, https://www.grin.com/document/1291753

Publish now - it's free

Comments