Financial Growth Mauritius

INTRODUCTION

The financial sector is the heart of any vital economy as it contributes to economic prosperity. That is also the case for Mauritius. It accounts for approximately 2% of total employment in the economy but contributes 10% of gross domestic product. The sources of investment are creating more demand, better delivery of services thus leading to a more competitive financial sector. Thus it becomes important to know by how much the financial indicators in the economy are affecting economic growth. This chapter will mainly summarise the steps taken in order to do this dissertation.

LITERATURE REVIEW

This chapter attempts to review the literature available on the subject, that is to review the relationship which exists between finance and economic growth. First of all, we will explain economic growth. Then, the different ways in which financial sector affects economic growth will be explained. This will be followed by an analysis of the real finance-growth nexus and afterwards it will be more about the debate on whether it is the structure of financial sector which affects economic growth or not. Finally, the other determinants of economic prosperity will be reviewed.

FINANCIAL SECTOR IN MAURITIUS

As for this chapter, it gives an insight of the actual financial sector in Mauritius. It consists of mostly banks, the stock market, the global business, insurance and pension. Each of these will be explained backed by data which will be analysing their performance in the economy.

METHODOLOGY

This chapter will mostly be explaining the different tests that will be performed in order to do the analysis of the effect of finance on growth. First of all, the variables which will be used in the model will be explained and their relative trend will be analysed. Eventually, the model will be designed and the different tests for model acceptance will be elaborated in order that the model is a reliable one. It must be noted that time series data will be used in our case.

ANALYSIS

This part of the dissertation consists of testing the data using the methodology in Chapter 4. Before the model becomes fit for analysis, various tests will have to be done namely multicollinearity, stationarity, and autocorrelation among others. When all these problems have been corrected, then the final regression can be done and thus the analysis will follow afterwards. This chapter will help us to quantify the effect that financial indicators have on economic growth.

CONCLUSION

This part will give a final note of the dissertation by taking into account the results obtained from the model. Recommendations will be given and this will help to have a clear picture of whether our financial sector is heading in the right direction or not.

CHAPTER 2: LITERATURE REVIEW

2.0 INTRODUCTION

This chapter will review the tremendous amount of literature available on the effect of finance on economic growth. First of all, the literature will be reviewed on economic growth, then on the functions of the financial sector which indirectly affect economic growth. Afterwards other factors affecting economic growth will also be taken into consideration.

2.1 ECONOMIC GROWTH

Economic growth is the rate of increase of a national income aggregate, usually gross domestic product (GDP), as measured in national accounts. Growth should preferably be measured on a per capita rather than a total basis in order to capture the effect on welfare. Achieving economic growth is the appropriate means of enhancing social welfare which is a principle objective of society and governments (Clayton and Radcliffe 1996; Cochrane and Shaw Bell 1956).

Kuznets (1968) stated that economic growth is the long term increase in the capacity to supply different economic goods and services. He advocated that it is calculated by comparing the change within GDP per capita between specified time periods. Economic growth is also viewed as increased stocks of capital goods. Solow and Swan (1950) developed a model which included labour-time, capital goods, output and investment in order to analyse long run economic growth. This is in line with neo-classical growth models suggesting that factors such as capital, labour and human capital should raise output levels over time but with diminishing returns in the long run.

There are various determinants of economic growth. Economic growth can be boosted persistently through exogenous factors such as improvement in technology, embodied in capital stock, human capital investments, research and development spillovers, economies of scale, among others. However, Jorgenson (1995, 2005) found that physical capital accumulation does not count for long run economic growth. Also it has been noticed that engines of growth vary in different countries. Montiel (1999) found that productivity accounted to 4.4% of growth in African countries while capital accumulation amounted to 64.4%. Similarly Young (1994) found that growth in South East Asia was mostly explained by labour force growth during 1966 to 1990.

Other factors affecting growth include culture, education, the role and size of government, geography, the extent of competition, macroeconomic conditions among others. Brenner (1998) questioned the fact that whether financial innovation, a change in company strategy or a change in government policy led to economic growth. For the scope of this dissertation, the role of the financial sector development will be highlighted towards its contribution to economic growth.

2.2 FINANCIAL SECTOR DEVELOPMENT

There is a tremendous amount of literature about the link between finance and growth. Before analyzing this relationship, we must have an insight of what a financial sector is, what its development means to society in general and what are the main functions of the financial sector.

A financial sector is all the wholesale, retail, formal and informal institutions in an economy offering the financial services to customers, businesses and other financial institutions. It consists of banks, stock exchanges, insurers, credit unions, microfinance institutions and money lenders.

It should be noted that the key features of a modern financial system are namely:

An effective banking system with an effective central bank

Efficient public finance and public debt management

Stable money in terms of low inflation and stable exchange rates

Securities markets to improve the access to ameliorate the access of governments and firms to finance capital

Firms that issue stocks and bonds to pool capital

Also, there are various ways in which a financial sector is said to develop namely with an increase in the amount of funds provided, efficiency of the legal framework, increase in the variety of financial institutions and services and better access of the population to the financial services.

The financial sector is thus the foundation of economic life, contributing to prosperity by channeling savings into new economic activities, enabling investment, controlling risk and easily transferring payment. Modern growth theory identifies two specific channels through which the financial sector might affect long run growth:

Through its impact on capital accumulation (human capital as well as physical capital)

Through its impact on the technological advancement

Merton and Bodie (1995) suggested that financial systems naturally affect the allocation of resources across space and time.

Levine (1997) found that the financial sector contributed to income growth by fulfilling five basic functions namely providing information about investment possibilities and allocation of capital, controlling firms and ensuring corporate governance after providing finance, management of risk and diversification, mobilizing savings and finally facilitating the exchange of goods and services.

A review of the literature will be done as to how each of these functions helps to influence savings and investment decisions, thus boosting economic growth. It must be noted that Merton and Bodie (1995) too identified five main functions but classified them differently. For the purpose of this study, Levine’s ideas will be reviewed. Eventually, the overall relationship between the financial sector and economic prosperity will be discussed.

2.2.A ACQUIRING INFORMATION ABOUT INVESTMENT OPPORTUNITIES & ALLOCATION OF CAPITAL

Ramakrishnan and Thakor (1984) analysed the role of acquiring information about the financial system, which they believed would enable efficient allocation of resources. They based their model on the fact that financial intermediaries encouraged provision of information and in turn, this information is sold to economic agents, namely savers.

Financial intermediaries can reduce information costs by obtaining and comparing information about many competing investment opportunities available for all their clients, thus allowing better allocation of capital (Boyd and Prescott 1986). The information costs may arise when assessing firms, market conditions prevailing and so on. If economic agents relied entirely on their own resources, it would not be possible for them to make a worthy investment. Therefore, groups of individuals may form financial intermediaries that decide to perform the costly process of researching investment possibilities for other people.

Greenwood and Jovanovic (1990) found that these intermediaries enabled the choice between investments depending upon their judgements about expected returns, thus putting aside the less worthy project. However, this implied that investors have good information about firms and market conditions (Bagehot 1873). The level of technology too can be increased as the entrepreneurs who are sure to succeed with the provision of new goods and production processes will be identified (King and Levine 1993).

The acquisition of information can be from either banks or stock markets. Grossman and Stiglitz (1980) found that the more liquid a market is, the better it is to benefit from this information and thus make a profit. The relationship between efficient markets, acquisition of information and economic prosperity has been analysed by Aghion and Howitt (1999).

As for the role of banks, Tsuru (2000) advocated that excessive proximity between banks and companies can reduce the cost of capital but at the cost of generating an unprofitable investment. This can negatively affect the ability of financial intermediaries to have a good investment selection based on information available. However, Sharpe (1990) noticed that where there information symmetry exists between two parties, then costs of information can be reduced, making the investment profitable.

2.2.B CONTROLLING FIRMS & ENSURING CORPORATE CONTROL

Financial intermediaries analyse and control the behavior of enterprises on behalf of the investors as the latter would not be in a position to do so. As a result, this encourages managers of the firms to perform well, for example managers would not be tempted to defraud the investors due to the presence of corporate control. This function affects both savings and allocation decisions, and eventually affecting growth. Bencivenga and Smith (1991) discovered that the facilities available to enhance corporate control tend to lead to faster capital accumulation and growth.

With constant monitoring of the manager’s performance, savers feel more willing to undertake investment and thus, it allows this to turn into fruitful and profitable arrangements (Stiglitz and Weiss 1983). However, Schleifer and Vishny (1997) observed that sometimes managers benefit at the expense of shareholders and society. This is due to the fact that market frictions in the financial sector prevent shareholders from exerting corporate control well.

One of these frictions was analysed by Grossman and Hart (1980) whereby it was found that large owners were more able to get information compared to other small shareholders. Jensen and Meckling (1976) found that this resulted to conflicts among them. Eventually, these conflicts affect growth adversely (Wolfenzon and Yeung 2005).

It must also be noted that sometimes shareholders are not apt to monitor the work of managers constantly. Also, they tend to believe the fact that other shareholders will be monitoring the work instead of them, therefore they end up not controlling manager’s job.

Levine (2005) made an in depth analysis of how liquid equity markets, debt contracts and banks help to improve corporate control and lead to growth. Jensen and Meckling (1976) analysed the importance of stock markets in enabling corporate governance. As for Aghion, Dewatripont and Rey (1999), they establish the relationship between debt contracts and growth.

2.2.C MANAGEMENT OF RISK & DIVERSIFICATION

Financial intermediaries can ensure that any individual can get back his money whenever he wants. This encourages not only risk diversification but also risk reduction. Returns increase and savings are encouraged, thus promoting economic growth. This part will be analysed by looking at risk diversification and risk sharing.

It is better to invest in a wide variety of projects where the expected returns are not correlated than to invest in a single project. This is due to the fact that savers do not prefer to face high risk situations. Financial intermediaries such as banks and stock exchanges facilitate risk diversification by allowing investment to be made in riskier projects with higher expected returns (Saint Paul 1992, Obstfeld 1994). This function can affect long economic growth by changing resource allocation and savings rates (Levine 2005). This can be explained by the fact that investments generating higher return tend to be riskier. Financial intermediaries thus allow people with to shift towards projects with higher return (Gurley and Shaw 1955).

Risk diversification can also have an impact on technological change. King and Levine (1993) found that by making more capital available to innovators, financial intermediaries which promote diversification also help to increase technological change. They termed it as “stimulating innovating activity.”

Another advantage is risk sharing. However, theory tends to focus on markets rather than intermediaries when explaining risk sharing. Risk-sharing is advantageous to borrowers and investors. Investors spread their investments across many firms while borrowers obtain finance for projects whose risk is borne by many. Investment diversification, insurance and hedging are examples of risk sharing. Allen and Gale (1997) examined the role of intermediaries in facilitating inter-temporal risk sharing. Devereux and Smith (1994) and Obstfeld (1994) show that greater risk sharing through internationally integrated stock market induces a portfolio shift in low return to high return investment, thereby accelerating growth. Similarly, Acemoglu and Zilibotti (1997) show that financial intermediaries encourage a reallocation of savings towards high return projects and this positively affects growth. However Fulghieri and Roveli (1998) argue that financial markets do not allow for risk sharing since different generations participate in the market at different times.

Another role which emerges from diversification is liquidity. Savers prefer to avoid risk and indirectly prefer liquidity. Levine (1991) found that financial intermediaries provide both functions, that is medium to long term capital for investment as well as liquidity for savers. However, since more liquidity makes it easier to dispose of shares, Shleifer and Vishny (1986) and Bhide (1993) argue that liquidity reduces the incentives of shareholders to undertake costly task of monitoring managers and this has a negative impact on growth. Levine (2005) found that financial intermediaries may increase liquidity of savers, at the same time reduce liquidity risk and eventually influence economic growth. Diamond and Dybvig’s (1983) model of liquidity highlighted two main points namely that savers have the possibility to choose between either illiquid, high return projects or liquid but low return projects. Banks offer both kinds of investments. However, by opting for an appropriate mix of liquid and illiquid investment, banks reduce liquidity risk that savers have to face and at the same time facilitate long run investments in high return projects. However, if agents prefer markets to banks, then banks will not be able to perform this role well (Jacklin 1987).

2.2.D MOBILISING SAVINGS

The provision of saving facilities enables people to store money in a safe place. It is these savings which are later used to allow people who need finance to benefit, thus encouraging capital accumulation investment through credit creation.

Levine (2005) observed that pooling of savings led to problems such as high transactions costs which arose when collecting savings and informational asymmetries which are necessary to make savers feel confident in letting their savings in the hands of another. Boyd and Smith (1992) too believe that in order to mobilise capital, “mobilisers” must convince savers that their investment is worthwhile. Pooling occurs through intermediaries, where investors give their wealth to intermediaries who will use these funds to invest in other firms (Sirri and Tufano 1995).

Acemoglu and Zilibotti (1997) show that pooling of savings from different people and later investing it in a diversified portfolio leads to a better allocation of investment and this has positive impact on economic growth. Along with the effect of good savings mobilizationon capital accumulation, mobilisation also enhances technological innovation, thereby encouraging growth. This has been propounded by Mc Kinnon (1973).

In Fry (1978), the results of pooled time series analysis for seven less developed countries support the view that financial conditions influence savings and growth. Finally, De Gregario (1996) found that credit may also be made available to finance investment in education or health, and can thus promote the accumulation of human capital. Thus, savings mobilisation can have a significant impact on growth by increasing investment, productivity and human capital.

2.2.E FACILITATE EXCHANGE OF GOODS & SERVICES

The financial sector enables transactions in an economy by providing a mechanism for clearing and settlement of payments and financial claims, as well as by reducing information and transaction costs. Thus, it allows greater specialization and productivity, which allows more technological innovation and growth.

Greenwood and Smith (1996) have modeled the connection between exchange, specialization and innovation and found that specialization increased the number of transactions, and in turn, financial intermediaries lowered transaction costs, which encouraged specialization. Productivity gains are thus encouraged. However, they also showed that the fall in transaction costs did not encourage better technology.

Smith (1976) noted that money played an important role in lowering transaction costs, thus encouraging technological innovation. King and Plosser (1986) found that lower transaction costs enabled a better medium of exchange for the transfer of goods and services.

2.3 FINANCE-GROWTH NEXUS

There is no dearth of literature concerning the relationship between finance and economic growth as this link has been extensively studied by researchers. An analysis of how each of the functions of the financial sector helps to promote economic growth has already been done. The relationship between overall financial development and economic growth will now be discussed.

The origin of finance led growth hypothesis can be traced back to Bagehot (1873). Schumpeter (1911) contends that the services provided by financial intermediaries are essential for growth. Mc Kinnon (1973) and Shaw (1973) analysed interest rate ceiling, direct credit program and high reserve requirements and concluded that a poorly functioning financial system may contribute to economic growth.

Levine (1991) explains how stock markets influence growth by improving firm efficiency and Bencivenga and Smith (1991) explain in their study that a well-functioning financial system would improve the level of investment towards non-liquid objects, which would be beneficial to the economy. Blackburn and Hung (1996) found that the role of the financial sector helps boosting economic growth by providing services, lowering transaction costs and by channeling greater savings towards new projects.

Atje and Jovanovic (1993) explained how financial system helped investors avoid risk and provided funds, thereby providing a good allocation of resources which is profitable to the economy.

To prove these finance led growth theories, there has been substantial body of empirical work which has been performed. Most of the models used were mostly cross-section studies, but later more panel and time series studies were used.

Goldsmith (1969) wanted to find whether finance caused growth or whether it was the structure of the financial sector which was more important. Thus data on 35 countries were compiled over the period 1860 to 1963 on the value of financial intermediary assets as a share of economic output. He arrived at the conclusion that financial development led to faster economic growth.

However, his work was not free of problems namely:

His research involved only 35 countries

It did not control for other factors influencing economic growth, therefore we will not know whether it is really finance or the other factors which influence growth

It does not examine whether finance is linked with productivity growth

The direction of causality could not be found, that is whether it was finance which caused growth or the other way round

Finally no conclusion could really be drawn concerning the fact whether financial structure matters for growth or not

King and Levine (1990) tried to overcome the drawbacks of Goldsmith’s work by:

Studying more countries, that is 77 countries over the period 1960 to 1989

Controlling for other factors affecting long run growth

Examining whether finance is linked to capital accumulation and productivity growth

Rousseau and Watchel (2000) used the dynamic panel study to examine the relationship between stock markets, banks and growth. Their results portray that both components play an important role in the economic growth an economy. Jung (1986) and Demetriades and Hussein (1996) made use of time series data to explain the finance-growth link. They used vector autoregressive procedures and found that finance caused growth. Another interesting point to note on their study is that the finance-growth relationship was bi-causal.

It was thus not only important to know whether finance led growth but it was equally important to know whether the other way round was true. Kar and Pentecost (2000) investigated the direction of causality between financial development and economic growth for the period 1963-1995. Their findings suggest that relationship depends on measures of financial development.

Rousseau and Watchel (1998) found one-way causality in the relationship between financial development and economic growth in the case of five OECD countries during the period 1871 to 1929. Kul and Khan (1999) found bi-directional causality for all countries in the sample.

However, it must be noted that the results obtained do not necessarily imply that finance is always exogenous to economic growth. In fact, some theorists even believe that this relationship is over-emphasized (Lucas 1988). Robinson (1952) famously stated that “where enterprise leads, finance follows.” Here, it implies that finance does not cause economic growth. Levine and Zervos (1998) argued in their study that higher returns and improved risk could encourage lower savings rate, which could lower economic growth.

Thus, the debate about the finance-growth relationship is far from over, despite the fact that most people will tend to be in favour of the argument that finance is in fact an important determinant of economic growth. Also, it has been observed that researchers have put emphasis on other factors which helped to promote growth. Some believe that the structure of the financial sector is important, while other believe it does not matter. Some also studied factors like inflation, trade, legal framework among others, and their contribution to economic development.

2.4 STRUCTURE OF FINANCIAL SECTOR AND GROWTH

There have been debates on whether the structure mattered when analyzing the finance-growth link. Various different views have been obtained. The structure of the financial sector implies whether it is market based or bank based and if it is both, whether one has a more prominent place than the other. Debates about the importance of bank based and market based financial system have always existed. In fact, both play an important in the financial sector of any country and thus contribute towards economic growth.

Many authors believe that banks are better when compared to markets, whereas others believe that markets are more beneficial. Gerschenkron (1962) argued that banks are better in their role of financing when compared to markets in under developed economies, and also claim that state owned banks could be helpful in case of market failures. Since banks have the power to make additional funding when a project is in progress, Stulz (2000) argued in favour of banks when compared to markets.

Furthermore, many authors stress on the drawbacks of markets when compared to banks. Stiglitz (1985) found that well-developed markets reveal information to investors and thus discourage them from investing in researching firms. This tends to stymie economic growth. Banks, on the other hand, do not reveal their information immediately and thus encourage investment in research companies, thus having a positive impact on growth.

Shleifer and Vishny (1997) found that markets cannot exert corporate control well enough as the problem of insider dealings exists. Furthermore, Grossman and Hart (1980) argued that rapid public dissemination of information reduces incentives to obtain information and gaining corporate control. Chakraborty and Ray (2004) conclude that banks resolve the problem of insider dealings when compared to markets.

On the other hand, the market-based view highlights the drawbacks of banks. Hellwig (1991) and Rajan (1992) conclude that banks prevent innovation by extracting informational rents and by protecting established firms. Also banks tend to be more prudent with their debt issuing strategies, thus investments are postponed and economic growth retarded (Morck and Nakamura 1999).

Further defects of banks as stated by Levine (2005) are that

Banks tend to rely on a ‘main bank’ and thus take much time to employ growth strategies

They tend to use more capital intensive processes

They tend to make lower profits

Also they might not be able to react efficiently to uncertain situations (Allen and Gale 1999). Bank based systems are not fully involved in corporate control over firms. Bankers will not necessarily work in the advantage of customers and society but will prefer to do so in their own benefit. Powerful banks may join with firm managers to act against investors (Hellwig 1998; Wenger and Kaserer 1998). They might also impede the flow about information about firms (Black and Moersch 1998).

Proponents of market based system believe that market will more effectively identify bankrupt firms and prevent them from having a negative effect on the economy. Banks would not have been able to do so. Finally, they also believe that state owned banks are more interested in achieving political goals when compared to markets which tend to be independent (La Porta, Lopez De Silanes and Shleifer 2001). However, instead of making in depth analysis at both systems, some theorists believe that financial arrangements are of no importance, instead it is the overall financial development which must be taken into consideration (Merton 1992; Merton and Bodie 1995; Levine 1997).

Financial arrangements arise to assess potential investment opportunities and to perform 5 distinct roles as defined by Levine (1997). By providing financial services, different financial systems promote economic growth to a greater or a lesser degree. Thus, it is not important whether financial system is more market based or bank based. The main issue is in fact the creation of an environment in which these two provide sound financial services.

Another criticism of giving importance to the mix of the financial system is the fact that both provide complementary services which contribute to economic growth of a country (Boyd and Smith, 1998). This is because banks help in enhancing liquidity and stock markets enables corporate control and by offering alternative means of financing investment, they tend to reduce the problems generated by banks.

Stulz (2000) argues that banks limit its exposure thereby reducing risk and thus, co-existence of both systems tend to increase competition among them, thus leading to reduction in transaction costs and increased investment, generating positive impact on economic growth.

Thus, the distinction between bank based and market based system is gradually fading out due to the fact that both systems complement each other (Scholtens, 1999). But still, it is hard to ignore all the relative advantages and disadvantages of both systems.

Furthermore, many other factors are believed to affect economic growth some of which are legal factors, cultural factors, trade liberalization, macroeconomic factors like inflation and political factors among others. These factors may affect economic development to a larger or lesser degree, but the fact remains that they cannot be ignored when analyzing the factors contributing to growth.

2.5 LEGAL FACTORS

It is a widely accepted fact that law and policies regarding financial sector matters to economic growth and stability. The ‘Law and Finance’ literature (La Porta et al 1988), asserts that a country’s legal framework is related to the ability to protect individual’s property rights, which in turn affects investment decisions, finally bearing an impact on the accumulation on capital to achieve financial deepening. Levine et al (2000) point out that legal and accounting reforms strengthen creditor rights and contract enforcement, thus accelerating economic growth.

Also, one of the basic functions of the financial systems, as stated by Levine was the corporate control, and corporate control would be inefficient without a sound regulatory framework. La Porta et al (1997, 1999) analysed the effect of legal system on financial system. Their work shows that legal rules on investor protection and the overall quality of legal system affect corporate behaviour. Bergloff and Von Thadden (1999) stress on the fact that countries need to improve their laws in order to improve corporate governance and thus achieve desirable economic effect, one of which is economic growth.

2.6 POLITICAL FACTORS

Political stability in any country helps a lot towards the sound health of an economy. Thus, political factors too tend to affect economic growth of a country. Two countries with the same financial sector may still be different due to the effectiveness of their governments and the political factors which prevail in the country. Roubini and Sala-I-Martin (1991, 1992) show that government intervention in financial market tends to stop the process of economic growth. Barro (1991) indicates that political instability is associated with negative growth outcome in a large cross section of countries.

Political instability is associated with uncertainty about the business to undertake, about the investment and financial decisions to make. This increases the level of risk which usually investors prefer to avoid in order to earn returns. Political instability tends to delay investments, thus having negative impact on economic growth. Leblang (2003, 2004) analysed the effect of political indicators on growth and also included the fact that election period may matter for economic stability.

Handreau and Zumer (2004) also argue that governments and frequent elections can bolster a nation’s reputation for fiscal responsibility, thus lowering interest rate and fostering economic growth through financial development.

2.7 INFLATION

Among the important determinants of economic growth is inflation. The relationship between inflation and economic growth is not necessarily a direct one. Indeed, it can be complex as empirical studies provided mix results about this relationship (Haslay 1997).

Inflation alters return and thus affects real value. Inflation is also related to financial repression and the channel by which it affects growth is through the financial sector. There are two distinct relationships between inflation rate and economic growth. Firstly, a higher level of inflation can accelerate growth by the creation of demand pressures. The second is the negative long term relationship which exists. Khan and Senhadj (2001) analysed the inflation and growth relationship for industrial and developing countries. They concluded that there was the existence of a threshold beyond which inflation exerts a negative effect on growth and inflation rate below the threshold level have no effect on growth. Rousseau and Watchel (2000) analysed the interaction between growth-finance and inflation-growth relationship. They found that direct relationship between inflation and growth exists but gradually fades if inflation is moderate.

Earlier studies fail to establish any meaningful relationship between inflation and economic growth (Tun Wai 1959). Kearney and Chowdhury (1997) analysed 70 countries, out of which 48 were developing countries over the period 1960-1989 and found no causal relationship between inflation and economic growth in 40 per cent of the countries. They also observed mixed responses, that is both positive relationship and negative relationship.

2.8 TRADE

Liberalising trade is a way of benefitting from specialization and scale economies. But if trade promotes financial development, this offers a more complex way in which it may raise economic growth. This means that the relationship is not necessarily a direct and simple one.

Trade and financial development may be linked, either because of political reasons or because of foreign competition which affects demand and thus affects economic growth. Rajan and Zingales (2003) show that goods market openness can improve finance, as it is in their interest to benefit from competition. Svaleryd and Vlachos (2002) believe that openness to trade is associated with greater risks, such as exposure to external demand. However it helps to create a new demand for external finance and thus positively affects economic growth.

CHAPTER 3: FINANCIAL SECTOR IN MAURITIUS

3.0 INTRODUCTION

Over the last decade, Mauritius has realised substantial progress in its financial sector. Capitalising on its strategic location and relying on its sound domestic economic base, Mauritius has asserted itself as a premier international business hub in the Indian Ocean region. Mauritius has succeeded in demarcating itself as a serious financial services jurisdiction, offering a combination of quality services, and reliability to investors through its innovative regulatory framework and the increasing value-added services offered on the island. The modernisation of the onshore financial sector through financial liberalisation, the efforts of the Government to widen the pool of institutional investors, the establishment of an offshore sector and the stock exchange have elevated the financial sector of Mauritius to the rank of engine of growth and significant contributor to national income.

3.1 BACKGROUND

During the past thirty years, the Mauritian economy has diversified from a sugar-cane monocrop economy in the 1970’s to one based on sugar, manufacturing which is composed of textiles and garments mainly, and tourism in the 1980’s. Global business and Freeport activities have also been growing continuously since the mid 1990s and now, Mauritius has become the leading center for offshore financial activities in Africa and is also great source of foreign direct investment.

Nowadays, the government is leaving no stone unturned in order to promote financial services as it has been noticed that it is a high growth and high value added sector which has been consistently growing at a rapid pace in recent years. The insurance sector is growing at an annual average rate of 5.5%, the banking industry is expanding at the rate of 7.5%, while the non-bank and the non-insurance financial services has been growing at an average rate of 12% during the past four years. In the Annual Report of the Financial Services Commission in 2003, it is stated that assets of contractual savings institutions, namely those of the insurance and pension funds industry have increased to Rs 59.2 billion representing 42% of GDP. Stock market capitalisation grew by 32.7% and market turnover increased by 5.2%, thus showing the importance of the financial sector in Mauritius in today’s time.

3.2 STRUCTURE OF MAURITIAN FINANCIAL SECTOR

The financial sector of Mauritius is such that it operates under two main wings namely the Bank of Mauritius (BOM) and the Financial Services Commission (FSC). The Bank of Mauritius regulates the operation of Banks and the FSC is a regulator of non-banking financial institutions, namely the Capital Market, Global Business, Insurance and Pension. Each of these will be further analysed:

3.2.A BANKING

The banking sector plays an important part in the financial sector of Mauritius and consists of the Bank of Mauritius, offshore banks and domestic banks. The banking sector comprises of over two-thirds of the domestic financial sector and has grown at an average of 13 percent over the last years.

3.2.A(1) BANK OF MAURITIUS

The Bank of Mauritius, also the Central Bank, was established in September 1967. The Central Bank is known as the bank of banks, as all banks have an account with the Central Bank. Thus it has been set up as the authority which is responsible for the formulation and execution of monetary policy consistent with stable price conditions.

The Bank of Mauritius has the responsibility of supervising:

Licensing of Banks

Capital Adequacy

Quality of Management

Liquidity of Control

Concentration of Risk

Role of External auditors

On-site Examination

Off-site surveillance

Control of Advertisements

Confidentiality of Information

Identity of Customers

The Central Bank of Mauritius has as main function to issue currency to allow it to be in circulation and the secondary objectives of the Central Bank is to act as a supervisor on the activities of all the other banks in the country.

It must be noted that various monetary policy measures have been taken since the Bank started operations in 1967. At first, most of the policies in Mauritius were expansionary. In 1972, however, there was the end of this type of policy which was prevailing since 1969. That is why the minimum cash ratio has been fluctuating since then. In the year 1967, the cash ratio was fixed at 5.5%. Since then there has been many ups and down, but since 2006, it has been maintained at a minimum of 4%. This encourages stability in the credit creation process of the banks.

3.2.A(2) DOMESTIC BANKS

Financial Institutions may be licensed by the Bank of Mauritius to transact domestic banking business. Presently, there are 19 domestic banks, which is an indication of how the banking sector is of vital importance in an economy.

Domestic banks accept various types of deposits from the public such as personal savings deposit, fixed-term deposit among others. They also grant loan, deal in foreign exchange, provide safekeeping facilities and perform various other services. Domestic banks are also engaged in the provision of leasing, stock broking, asset and fund management, investment and private banking business, insurance agency, portfolio as well as custodial management.

The Mauritius Commercial Bank and the State Bank of Mauritius control approximately 70% of the systems asset, thus there is high concentration of domestic resources in the hands of very few banks.

The banking sector of Mauritius is one which is continuously innovating in order to attract more and more clients. One way to prove this is the increase in electronic banking transactions. This can be shown as follows:

Table 3.1: Electronic Banking Transactions

	DEC 04	DEC 05	DEC O6	DEC 07
No. Of ATM in operation	283	313	326	373
No. Of Transactions	3285091	3698436	3784838	4496145

Source: Bank of Mauritius Monthly Statistical Bulletin March 2008

The table above clearly shows how banks are important in an economy and how they are efficient. In a lapse of time of 3 years the number of ATM has seen a rise of approximately 30 percent. As for the number of transactions which have occurred, this too has increased by 38 percent. This shows that people in Mauritius rely on banks in order to save their money.

3.2.A(3) OFFSHORE BANKS

Mauritius offers an ideal environment for foreign banks and other financial institutions to conduct their international transaction. Offshore banks are licensed to conduct banking business or investment banking business in currencies other than the Mauritian rupee. They undertake deposit-taking, trade financing, fund management, investment advisory services and trusteeship of offshore trusts. Currently there are 11 offshore banks in Mauritius and this sector is experiencing a sustained growth.

3.2.B CAPITAL MARKET

The capital market is the market for securities, where companies as well as government can raise long-term funds. The capital market of Mauritius consists of the Stock Exchange of Mauritius (SEM), the Development and Enterprise Market (DEM), Leasing companies among others.

3.2.B(1) THE STOCK EXCHANGE OF MAURITIUS

The Stock Exchange of Mauritius was incorporated in Mauritius in 1989 as a private limited company. It started its operation with 5 listed companies and a market capitalization of US$ 92 million. Nowadays, there are 41 companies listed on the Official Market and it consists of a market capitalization of around US$ 1670.84 million. Also, companies listed on the SEM are classified into sectors such as commerce, industry, investment, leisure & hotels, sugar, transport, foreign and finally banks & insurance and other finance. There are three market indices namely the SEMDEX, the SEMTRI and the SEM-7.

The SEMDEX is an index of prices of all listed shares and it is listed by market capitalization. As for the SEM Total Return Index (SEMTRI), it is an index to provide market’s evolution over time as it captures the price movements of listed stocks but also provides a good measure of total return. The SEM-7 is an index which measures movements in the 7 largest eligible shares on the Official List in terms of capitalization, liquidity and investibility. All 3 indexes are following an upward trend and this is a good sign for the economy of Mauritius.

Table 3.2: Selected Market Indicators

YEAR	MKTCAP/GDP (%)	TURNOVER/MKTCAP (%)	TURNOVER/GDP (%)
1989	4.32	0.97	0.04
1990	9.66	2.35	0.23
1991	10.97	1.67	0.18
1992	13.30	2.41	0.32
1993	26.39	4.64	1.22
1994	44.55	5.33	2.37
1995	39.58	4.39	1.74
1996	42.20	4.80	2.03
1997	41.92	8.11	3.40
1998	45.53	5.64	2.57
1999	38.73	4.74	1.84
2000	31.10	5.52	1.72
2001	24.33	10.24	2.49
2002	26.99	4.46	1.20
2003	32.55	5.83	1.90
2004	38.19	4.21	1.61
2005	43.18	5.68	2.45
2006	56.96	5.12	2.92

Source: SEM Factbook 2007

From the table above, we can analyze the trend of the given market indicators since the time the Stock Exchange was set up that is in 1989 up to the year 2006. From the table, it can be seen that in 1989, market capitalization was only 4.32% of the Gross Domestic Product and since then, the share of market capitalization rose continuously until the year 1998, and afterwards it fell until the year 2002 and eventually rose until the year 2006. Up to now, market capitalization forms part of 56.96% of GDP which is quite a considerable amount. This shows that the Stock Exchange is contributing a lot to the national income of the country.

As for turnover, it does not contribute much to market capitalization, starting from 0.97% in 1989, it fluctuated both upwards and downwards but within a same range and arrived at its peak in 2001, and eventually fell until the year 2006. Turnover does not contribute much also to national income. However, a rise can be noticed from the year 1989 where it contributed only 0.04% of GDP and in the year 2006, it contributes 2.92% of GDP, which is minimal but which has experienced an increase.

3.2.B(2) THE DEVELOPMENT & ENTERPRISE MARKET

The Development & Enterprise Market (DEM) is a market set up on August 2006 and is designed for Small and Medium-sized Enterprises and newly set-up companies which possess a sound business plan and demonstrate a good growth potential. The DEM consists of 53 companies up to now and since it has been launched, the Over-the-Counter market of the SEM was phased out at the end of 2007.

Table 3.3: DEM Foreign Investors 2006

MONTH	VOLUME	PURCHASES(Rs)	VOLUME	SALES(Rs)	NET PURCHASES
Aug 2006	920,288	35,703,291.30	20,838	665,103.2	35,038,188
Sept 2006	31,983	669,764.20	180,436	1,810,820	-1,141,056
Oct 2006	372,252	6,772,084.40	5,605	332,990	6,439,094
Nov 2006	87,891	2,611,150	6,715	269,717	2,341,433
Dec 2006	2,679,951	50,134,621.50	2,194	108,565	50,026,057
TOTAL	4,092,365	95,890,911.4	215,788	3,187,195.2	92,703,716

Source: SEM Factbook 2007

From the table, it can be seen that from the date the DEM was launched, it fell and remained low until November 2006. This may be due to the fact that the market was not well stabilized. However, in December 2006, net purchases experienced a considerable increase from Rs 2,341,433 to Rs 50,026,057, thus the DEM is slowly marking its territory in the financial sector of Mauritius.

3.2.B(3) LEASING COMPANIES

Leasing companies provide leases for equipment, machinery, plant, motor vehicles, tools and other assets to industrial, agricultural, commercial, service sectors and individuals.

The list of leasing companies licensed by the Financial Services Commission as at 30 June 2007 is as follows:

SICOM Financial Services Ltd

La Prudence Leasing Finance Co. Ltd

Capital Leasing Ltd

Oceor Lease Mascareignes Ltee

SBM Lease Ltd

The Mauritius Leasing Company Limited

In order for any person to conduct leasing business, he must be licensed by the Financial Services Commission., that is why the FSC has put in place a supervisory framework to better analyse the leasing companies which are not involved in deposit taking activities.

3.2.C GLOBAL BUSINESS

Offshore business activities became a significant sector in Mauritius as from 1992, when the Mauritian Offshore Business Activities Act came into force, which was active and effective in structuring offshore regimes in various sectors. International investors who wish to choose Mauritian vehicles can choose between two types of companies holding global Business License, either Category 1 (GBL1) or Category 2 (GBL2) for qualified global business activities.

GBL1 companies are allowed to carry out the following qualified business activities namely aircraft financing, leasing, asset management, consultancy services, employment services, financial services, funds management, information and communication technology services, insurance, licensing, franchising, pension funds among others.

GBL2 companies cannot be used for banking, insurance, or fund-related activities. These companies cannot transact business in Mauritian rupees with Mauritian residents. They cannot be public companies ant they must essentially have a registered agent who will ensure that the company complies with statutory requirements and communicate with the Mauritian authorities.

Table 3.4: Financial Results of Management Companies (2004-2006)

	2004	2005	2006
No of Management companies	75	68	67
Turnover (US$ m)	39,940	48,646	64,302
Profit Before Tax (US$ m)	12,214	21,042	28,328

Source: FSC Annual Report 2007

Despite the fact that the number of management companies in operation in Mauritius as at 2006 has fallen from 75 to 67, an increase of approximately 60% can be noted in the amount of turnover from the year 2004 to 2006. It must also be noted that the FSC has registered 84 management companies as at June 2007, thus showing the expansion of the global business. The profit before tax too follows the same trend thus showing that these companies are very profitable and have carved a niche in the financial sector of Mauritius.

3.2.D INSURANCE

The insurance sector also plays a key role in the financial sector of Mauritius. In order to analyse the trend in this sector, the change in assets which has taken place in long term insurance business from the year 2002 to 2006 will be analysed for the 13 insurance companies in Mauritius.

Table3.5: Assets of Insurance Companies

	2002 (%)	2003 (%)	2004 (%)	2005 (%)	2006 (%)
Mortgage loans	21	18	15	14	12
Other loans	4	2	3	3	2
Govt Securities	13	13	23	24	24
Shares & Debentures	38	48	39	32	36
Land & Property	4	4	4	4	3
Deposits & Securities	9	7	8	14	16
Other assets	11	8	8	9	7
Total(Rs 000)	27160257	32592002	37036429	42139645	50706193

Source: FSC Annual Report 2007

From the table above, it can be seen that the distribution of deposits and other securities and government securities in insurance companies have risen while the others have fallen. Despite this, the total assets figure has been very high and experienced a change of 87 percent from 2002 to 2006, thus showing how profitable this section of the financial sector is in Mauritius.

Not only that, gross premium rose by 9.34% while net premium grew by 10.2%. Overall retention by general insurers remained almost constant 51% of gross premium. It must also be noted that in 2006, the life market grew by 7.7%. this was considered as the highest rate since 2000.

3.2.E PENSION

Mauritius has a pension system which comprises of the Basic Retirement Pension, the National Pension Fund, the National Savings Fund and the Civil Service Pension Scheme. In order to know the projected benefits of this part of the financial sector, the Government Actuary’s Department in London made a review of the situation in Mauritius and made predictions for the future. They found:

Table:3.6: Estimated expenditure on contributory pension

(Rs Million)

YR ENDING 30 JUNE	2000	2005	2010	2015	2020	2025	2030	2035	2040
TOTAL CONTRIBUTORY PENSION	266	504	812	1232	1723	2309	2908	3473	4143
RETIREMENT PENSION MINIMUM GUARANTEE	22	18	23	26	31	33	31	25	30

Source: Govt Actuary’s Dept Review, London ,June 2000

The diagram shows that that total contributory pension is expected to rise tremendously by 2040. This study also showed that this expenditure as a percentage of GDP will rise from 3.4% to 4.3%.

Also, funds of managed pension schemes amounted to almost Rs. 11 billion as Dec 2006. This sector is profitable and is expected to grow further in the future.

3.3 CONCLUSION

As we have seen, the Mauritian financial sector mainly comprises of banks, stock market, global business, insurance and pension. Each of these is progressing in its own way and pace. But what is important to know is the fact that they are highly profitable in general, despite a higher concentration of banks when compared to others. That is why the financial sector comprises of 10% of the GDP which is a considerable amount.

CHAPTER 4: DATA & METHODOLOGY

4.0 INTRODUCTION

This chapter gives details about the variables that will be used in a model designed to quantify the effect of financial development on economic growth, that is to explain by how much the indicators of financial development affect economic growth.

First of all, we will proceed by explaining the dependant as well as explanatory variables that will be used and their respective trend will be analysed over the years 1986 to 2005.

4.1 DATA SELECTION

Data selection is of great importance in assessing the quality of the tests undertaken. In this case, time series data for the period 1986-2005 will be collected. Data like financial development indicators and economic growth indicators will be used.. Then the indicators of financial development will be analysed over the given period. Secondary data will be most appropriate for the types of tests that will be undertaken. They are more easily collected than primary data and they are readily available from many reliable sources. Data will be mostly available from:

Bank of Mauritius Annual Report – Banking data

Stock Exchange of Mauritius Fact Book – Stock Market data

Financial Services Commission – Non-Bank Financial data

Central Statistics Office Report – Economic Growth Indicators

4.2 DATA ANALYSIS

4.2.A INDICATOR OF ECONOMIC GROWTH

King and Levine (1993a, b) used four measures of economic growth namely:

Investment as a percentage of Gross Domestic Product

A proxy of productivity improvements

Average growth rate of capital

Average growth rate of per capita real GDP

For the purpose of this dissertation, the proxy for economic growth will be average growth rate of per capita real Gross Domestic Product (RGDP) as it takes into account the element of population as well as that of inflation.

Fig 4.1: Average Growth Rate of Real GDP per Capita

Source: Central Statistics Office Report

The average growth of real GDP per Capita will be the dependent variable in the study. From the diagram above, this rate has experienced wide fluctuations from the year 1986 to 2005.

4.2.B INDICATORS OF FINANCIAL DEVELOPMENT

There are many financial indicators which can be used to test whether they affect economic growth or not. For the purpose of this dissertation, 2 indicators of banking sector and 2 indicators of the stock market will be used as proxies for financial development. They are:

Ratio of Money to Income (RM2).

Ratio of Liquid Liabilities to Income (RLIQUID).

Ratio of Market Capitalisation to Income (RMKTCAP).

Ratio of Market Turnover to Income (RMKTT).

RATIO OF MONEY TO INCOME

The most commonly used measure of financial development is a ratio of some broad measure of money stock, usually M2, to the level of national income (King and Levine, 1993a, 1993b). It in fact measures the degree of monetization in a financial world.

Money stock refers to the total amount of money held by the nonbank public at a point in time in an economy. There are several ways to measure such an amount (called a monetary aggregate), but each includes currency in circulation plus demand deposits. M2 consists of currency with the public added to time and demand deposits. However it does not include foreign currency deposits.

There are various views about the use of these indicators in order to assess its effect on economic growth. Gregorio and Gudotti (1995) believe that a high level of monetization (M1/GDP) is mostly the result of financial underdevelopment, that is why they suggest the use of M3 or M2/GDP as a proxy for financial development. For the sake of this dissertation, the ratio of M2 to GDP will be used, and since it is important to have it in real values in order to take into account the inflation prevailing in the country, the real ratio of M2 to GDP will be used thus RM2.

Fig 4.2: Money Stock to GDP

Source: Bank of Mauritius Annual Report(Various Issues)

From the diagram above, it can be seen that the real ratio of money stock to income has risen constantly over the years 1986 to 2005. This implies that money supply has experienced more than proportionate change when compared to the change in gross domestic product and it has risen in order to meet the relative demand of money in the country.

RATIO OF LIQUID LIABILITIES TO INCOME

Another proxy that can be used is the ratio of liquid liabilities to income (Demetriades and Hussein 1993). This will help to measure the size of financial intermediation. Liquid liabilities include currency as well as liabilities of banks and non bank financial intermediation.

In principle, a rising ratio of broad money to income may reflect the more extensive use of currency rather than the increase in the increase of volume bank deposits. Therefore to obtain a more representative measure of financial development, it is better to use the ratio of banking deposit liabilities to income. For the sake of this dissertation, we will adjust the ratio of liquid liabilities to GDP with inflation, thus we will use RLIQUID.

Fig 4.3: Liquid Liabilities to GDP

Source: Bank of Mauritius Annual Report(Various Issues)

From the diagram above, this ratio has experienced a rise from the year 1986 to 2001, then followed a slight decrease and finally remained stable until the year 2005. However, liquid liabilities is only an indicator of size and might not necessarily indicate the real provision of financial services in an economy.

RATIO OF MARKET CAPITALISATION TO INCOME

Stock market capitalisation attempts to capture the size of the Stock Exchange of Mauritius. It represents the aggregate value of a company or stock. It is obtained by multiplying the number of shares outstanding by their current price per share. This value as a percentage of the gross domestic product helps to know the relative importance of the Stock Exchange in the Mauritian Financial economy. This value too has been adjusted for inflation, that is why the real ratio of market capitalisation to GDP will be used, thus RMKTCAP.

Fig 4.4: Market Capitalisation to GDP

Source: The Stock Exchange of Mauritius Factbook 2007

The diagram above shows that this ratio has followed an upward trend since the setting up of the Stock Exchange in 1989. However it started experiencing a fall and was at its lowest point in the year 2001. Afterwards it once again followed an ascension up to the year 2005, showing that the value of the shares on the SEM rose.

RATIO OF MARKET TURNOVER TO INCOME

The market turnover ratio measures the volume of domestic shares traded relative to the size of the market. It is given by the value traded on the SEM divided by the value of listed shares on the SEM. Since it acts as an indicator of low transaction costs, it can be used as a proxy for financial sector development. This value too will be adjusted for inflation, thus the real ratio of market turnover to GDP will be used (RMKTT).

Fig 4.5: Market Turnover to GDP

Source: The Stock Exchange of Mauritius Factbook 2007

From the diagram, market turnover to GDP has followed a sharp rise from 1992 to 1994, but later fell thus showing higher transaction costs in the stock market. Later, this ratio stabilised and followed this trend until the year 2005.

4.2.C OTHER FACTORS AFFECTING ECONOMIC GROWTH

These 4 variables above would act as proxies of financial indicators for the analysis of economic growth in Mauritius. However, there are other factors which are non financial indicators that affect economic growth. In the case of this study, the current price index (CPI) and the Gross Domestic Fixed Capital Formation (GDFCF) will be added to the other variables in order to develop a model which will analyse the variables’ effect on economic growth.

The trend of both these factors will now be analysed:

CURRENT PRICE INDEX

Fig 4.6: Current Price Index with Base Year 1986

Source: Central Statistics Office Report

A consumer price index (CPI) is an index number measuring the average price of consumer goods and services purchased by households. The percent change in the CPI is a measure of inflation. The CPI is, along with the population census and the National Income and Product Accounts, one of the most closely watched national economic statistics. The diagram shows the trend of CPI in Mauritius from the year 1986 to 2005. It started with fluctuations and eventually started to stabilise after 2001 and upto 2005.

GROSS DOMESTIC FIXED CAPITAL FORMATION (GDFCF)

Fig 4.7: GDFCF to GDP

Source: Central Statistics Office Report

Another factor which affects economic growth is investment. This is termed as the GDFCF and is defined as the total value of additions to fixed assets by resident producer enterprises, less disposals of fixed assets during the quarter or year, plus additions to the value of non-produced assets. In order to adjust for inflation, the real ratio of GDFCF to GDP will be preferred thus RGDFCF. Despite fluctuations, this ratio has remained high and continued to do so until the year 2005, thus showing that investment has seen an ascending trend in Mauritius.

4.3 METHODOLOGY

4.3.A ECONOMETRIC MODEL

In order to analyse the effect of financial indicators on economic growth, a time series model will be used, whereby data will be collected for the period 1986 to 2005. Normally it is assumed that economic growth is the explained variable and financial sector development is the explanatory variable.

Therefore the model can be written as follows:

RGDP = f (FD, X);

RGDP implies rate of real gross domestic product per head

FD implies financial development indicators and are the explanatory variables

X implies the other factors affecting economic growth

Therefore the model becomes:

RGDP=f(liquid liabilities, money supply, market capitalisation, market turnover, gross domestic fixed capital formation, CPI)

RGDP = α + β₁RM2 + β₂RLIQUID + β₃RMKTCAP + β₄RMKTT + β₅RGDFCF + β₆CPI + µ;

When the regression will be run, β_1,β_2,β_3,β_4,β₅are expected to be positive as they help to boost economic growth. On the other hand, β₆is likely to be negative because it affects economic growth inversely. However, deviations might occur and the predicted result does not actually occur. As for µ, it is the error term in the model andα is a constant.

It must be noted that the regression will be run using the Microfit Software 4.0 and the original results will be found in the Appendix.

4.3.B TESTS FOR MODEL ACCEPTANCE

P-Value

Adjusted Coefficient of Determination (R²)

Test for Multicollinearity

Test for Stationarity

Test for Autocorrelation

1. P-Value

The p-value obtained from regressing variables actually denotes the statistical significance of the model. The p-value is actually the probability value and is also known as the exact level of significance. It is the lowest significance level at which a null hypothesis can be rejected. This value can be compared with a fixed level of significance, which can either be 1%, 5% or 10%. The lower the fixed level of significance, the better. Thus, if the value is less than the fixed level of significance, it means that the model is significant. In our case, we will compare the p-value with 5%.

2. Adjusted Coefficient of Determination (R²)

When the regression is done, it becomes important to consider the goodness of fit of the fitted regression of line. The coefficient of determination R² is a summary measure that tells how well the sample regression line fits the data. It in fact gives the percentage of the total variation in Y explained by the regresson model. Since many regressions will be done in order to get the final result, the R² values will have to be compared. Then R² will no longer be used as it is necessary to take into consideration the number of X variables in the model. This is why it is better to use the adjusted R² instead of R² as it is adjusted for the degrees of freedom.

3. Test for Multicollinearity

Multicollinearity is the existence of an exact linear relationship among all the explanatory variables in a regression model. There are various reasons explaining why multicollinearity exists. These are mainly because of constraints on the model or in the population being sampled, the data collection methods employed, model specification and an overdetermined model. The main reason why multicollinearity exists mostly in time series data is due to the fact that the regressors share a common trend, thus leading to high correlation among them. Correlation of the regressors will be done on the Microfit Software itself. If the value is greater than 0.8, then it implies high correlation among the variables. In order to remedy for this situation, the variables can either be transformed by using the first difference. This will be explained in Chapter 5.3.B(4). Also, there can be the use of ratio transformation. Multicollinearity can also be avoided by dropping one variable from the highly correlated pair.

4.. Test for Stationarity

Time series data have to be stationary in order to have better forecasts. Non-stationary time series have a time-varying mean or a time-varying variance, or both. This is a problem because time series data will be studied only for the period under consideration, thus making it difficult to generalise to other time periods.

If non stationarity is suspected in a regression model, firstly the data has to be tested. This can be done by the Augmented Dickey Fuller Test (ADF). This test is done by adding lagged values of the dependent variable.

The ADF test is run on equations of the following type:

ΔRGDP_t = β₀ + β₁RGDP_t + ∑ β_2iΔ RGDP_t-1 + µ_t(i)

ΔRGDP_t = β₀ + β₁T + β₂RGDP_t + ∑ β_3iΔ RGDP_t-1 + µ_t(ii)

Where ΔRGDP_t is the first difference, t stands for the time, and k is the number of lagged values taken. This will be determined by the Akaike Info. Criterion (AIC) and the Schwarz Bayesian Criterion (SBC). The higher of both values will be used to find the t-statistic of the variable. Equation (i) is the ADF test which includes an intercept but not a trend whereas the equation (ii) is which includes an intercept as well as a trend.

The logarithmic form can also be applied in case the ADF of the level form is not stationary. This will occur when the critical ADF statistic at 5% is less than the T-statistic.

In case the logarithm form is applied, then the ADF test is run on:

LNRGDP_t = β₀ + β₁LNRGDP_t + ∑ β_2iΔ LNRGDP_t-1 + µ_t(i)

LNRGDP_t = β₀ + β₁T + β₂LNRGDP_t + ∑ β_3iΔ LNRGDP_t-1 + µ_t(ii)

It must be noted that the ADF test is done on the Microfit Software itself.

5. Test for Autocorrelation

Autocorrelation occurs in case the errors in the model are correlated. The classical model assumes that there is equal error variance. However, this is not always the case. There are many reasons why autocorrelation occurs namely inertia, specification bias, lags, manipulation of data, among others. First of all, autocorrelation has to be tested. This can be done by the Graphical method, the runs test, the Durbin Watson test or the Breush Godfrey test. The most popular method of testing for autocorrelation is the Durbin Watson test. The d statistic is the ratio of the sum of squared differences in successive residuals to the residual sum of squares (RSS). The value of the d test should converge to 2. Then it implies that there is no serial correlation in the model.

In case autocorrelation exists, then the Cochrane Orcut method is used to correct it. It is in fact an iterative procedure and it can be used to estimate not only an AR(1) scheme, but also higher order schemes. The Cochrane Orcut procedure is done on the microfit software itself.

4.4 WEAKNESS

The problems encountered in this study were that data for the stock market was not available for the year 1986 to 1988 since the Stock Exchange was set up in the year 1989. Also, data about the variables used in the model were already available from various sources but they were not adjusted for inflation. In order to have better predictions, the ratios had to be accounted for inflation in order that they become real values. Despite these lacunas though, all tests have been carried out objectively according to econometric principles.

4.5 CONCLUSION

As we have seen, once we have decided what variables will be used, the model is designed and eventually, various tests like that of multicollinearity, stationarity and autocorrelation need to be performed in order to have a model which is not prone to errors and that will enable us to have a good analysis.

CHAPTER 5: ANALYSIS

5.0 INTRODUCTION

This chapter deals with the analysis of the time series data taken from 1986 to 2005 in Mauritius, using a methodology which has already been elaborated in Chapter 5. Thus, various tests will be done including stationarity tests, multicollinearity and autocorrelation tests, as well as the tests for model acceptance. It must be noted that the explanatory variables are RM2, RLIQUID, RMKTT, RMKTCAP, RGDFCF, CPI and the dependant variable is RGDP, all of which have also already been explained in Chapter 5. The regression and the other tests will be performed using the Microfit Software 4.0. All original Microfit results will be found in the Appendix section.

5.1 MODEL 1

In this section, a regression over the period 1986-2005 will be carried out using econometric analysis in order to quantify the effect of the financial indicators on economic growth. Our original model is as follows:

RGDP = α + β₁RM2 + β₂RLIQUID + β₃RMKTCAP + β₄RMKTT + β₅RGDFCF + β₆CPI

From the given equation, the regression is run from the Microfit Software and the results are as follows:

Table 5.1: Ordinary Least Square Estimation

REGRESSORS	COEFFICIENT	PROBABILITY
α	0.91519	0.024
RM2	-0.00554	0.983
RLIQUID	-0.34183	0.381
RMKTCAP	-0.07372	0.765
RMKTT	-0.31799	0.927
RGDFCF	1.3078	0.018
CPI	-0.00754	0.032

R²	0.39676
DW-Statistic	2.3219

Source: Computed

Dependent variable is RGDP

20 observations from 1986 to 2005

Therefore our equation becomes:

RGDP = 0.91519 – 0.00554RM2 – 0.34183RLIQUID – 0.07372RMKTCAP – 0.31799RMKTT + 1.3078RGDFCF – 0.00754CPI

From the results above, we can observe that the p-value for most variables are not significant as they are less than 5% (RM2 98.3%, RLIQUID 38.1%, RMKTCAP 76.5%, RMKTT 92.7%). However, RGDFCF and CPI have p-values less than 5%, thus making them significant.

Before analysing further, we must note that we have not yet tested for multicollinearity, stationarity and autocorrelation. Since the regression results show that most variables are insignificant, we suspect that the variables might be facing problems of multicollinearity, stationarity or even autocorrelation. That is why the next step will be to perform these tests. Other problems with this original model is that its explanatory power is minimal and is at 39% approximately, once again showing that we cannot rely upon this model as its explanatory power is low.

5.1.A TESTING FOR MULTICOLLINEARITY

Testing for multicollinearity is important as it helps to know whether there is a strong correlation amongst the explanatory variables in a regression model. Correlations around 0.8 and 0.9 are not desirable as they will indicate multicollinearity. In such a case, the highly correlated variable has to be removed. The correlation among the variables is shown as follows:

Table 5.2: Correlation of the variables

	RM2	RLIQUID	RMKTT	RMKTCAP	RGDFCF	CPI
RM2	1.0000	0.87645	0.73423	0.80574	0.30817	-0.16062
RLIQUID	0.87645	1.0000	0.75897	0.79855	0.56539	-0.15989
RMKTT	0.73423	0.75897	1.0000	0.91124	0.26137	-0.19479
RMKTCAP	0.80574	0.79855	0.91124	1.0000	0.39563	-0.17188
RGDFCF	0.30817	0.56539	0.26137	0.39563	1.0000	-0.00635
CPI	-0.16062	-0.15989	-0.15989	-0.17188	-0.0063570	1.0000

Source: Computed

The results above show that there is highest correlation between DRMKTT and DRMKTCAP at 91%. Both these variables are proxies used for the stock market. The other values mostly are less than 0.8, therefore we will remove MKTT from the model in order that there is no multicollinearity. It can be noted that when correlation has been checked among the remaining variables, values were less than 0.8, thus eliminating multicollinearity in the model.

5.1.B STATIONARITY TEST

Before doing the regression, all the variables including the dependent variable must be checked for stationarity. This can be done by using the Augmented Dickey Fuller Test. Lagged values used must be specified and this will be done by using the AIC and SBC Criterion by using large lags at first and then reducing it by a statistical criterion. In this case, 7 lags have been chosen and the maximum of the AIC and the SBC criterion will be used.

It must be noted that it is assumed each variable is a random walk with a drift but no trend where:

H₀: Time Series are not stationary

H₁: Time Series are stationary

The ADF results are shown as follows:

Table 5.3: ADF Test on Level Form Variables

VARIABLES	T-STATISTICS	CRITICAL ADF STATISTICS(5% Level)
RGDP	-2.9046	-3.1485
RM2	1.5218	-3.1485
RLIQUID	-2.1703	-3.1485
RMKTT	-4.3995	-3.1485
RMKTCAP	-2.9140	-3.1485
RGDFCF	-1.8349	-3.1485
CPI	-1.4189	-3.1485
U	-4.2411	-3.1485

Source: Computed

(Original results found in Appendix)

In order to know whether variables are stationary or not, the T-statistic has to be compared with the Critical ADF statistic. If the value of the critical ADF statistic is greater than the T-statistic, then the variable will be stationary. From the results above, it can be noticed that only the variable RMKTT and the error term U are stationary. All other variables are not stationary as the T-statistic is greater than the critical ADF statistic. In this case, the variables have to be transformed in order that they become stationary.

This can be done either by applying logarithm to all the variables or using the method of first-difference. However, when applying logarithm to all variables, this has not solved the problem of non stationarity as all the T-statistics were greater than the critical ADF tests. Therefore, the first-difference method has been used. The ADF tests are calculated in the same way as it was done on the level form variables.

Thus, the ADF test on the first difference data is shown as follows:

Table 5.4: ADF Test on First Difference Data

VARIABLES	T-STATISTICS	CRITICAL ADF STATISTICS(5% Level)
DRGDP	-4.7116	-3.1803
DRM2	-3.2089	-3.1803
DRLIQUID	-4.4034	-3.1803
DRMKTT	-5.1743	-3.1803
DRMKTCAP	-9.6225	-3.1803
DRGDFCF	-3.9597	-3.1803
DCPI	0.9362	-3.1803

Source: Computed

(Original results found in Appendix)

Where;

DRGDP = First Difference of real growth rate in GDP

DRM2 = First Difference of real ratio of Money Supply to GDP

DRLIQUID = First Difference of the real ratio of Liquid Liabilities to GDP

DRMKTT = First Difference of real ratio of Market Turnover to GDP

DRMKTCAP = First Difference of real ratio of Market Capitalisation to GDP

DRGDFCF = First Difference of real ratio of Investment to GDP

DCPI = First Difference of the Current Price Index

From the results above, the critical ADF statistic is greater than the T-statistic in the case of DRGDP, DRM2, DRLIQUID, DRMKTCAP, DMKTT AND DRGDFCF. However, this is not the case for DCPI. This variable is not stationary even on first differentiation of the data. Thus, the variable CPI will be excluded due to nonstationarity. Furthermore, all the other variables have already been adjusted for inflation. It must be noted that despite th fact that DMKTT is stationary, it has already been dropped out due to multicollinearity problem.

The problem of stationarity has been solved by using the First-Difference of the data. Thus, our new model will be:

DRGDP = α + β₁DRM2 + β₂DRLIQUID + β₃DRMKTCAP + β₄DRGDFCF

The regression is run and the results for the new model will be shown in the next table. It must however be noted that we have not yet performed the test for autocorrelation.

Table 5.5: Ordinary Least Square Estimation

REGRESSORS	COEFFICIENT	PROBABILITY
Α	-0.028313	0.352
DRM2	0.39295	0.640
DRLIQUID	0.20558	0.798
DRMKTCAP	0.32757	0.394
DRGDFCF	1.4161	0.102

R²	0.33376
DW-Statistic	3.4261

Source: Computed

Dependent variable is DRGDP

19 observations from 1987 to 2005

Therefore, our equation becomes:

DRGDP = - 0.02831 + 0.39295DRM2 + 0.20558DRLIQUID + 0.32757DRMKTCAP + 1.4161DRGDFCF

The results above however, show that the regressors are not significant at 5% level. The probability of DRM2 is 35.2%, DRLIQUID is 64%, DRMKTCAP IS 79.8% and DRGDFCF is 10%, all of which are greater than 5%, thus making them insignificant.

Thus, despite having checked for multicollinearity and corrected for stationarity, this model cannot be used to analyse the effect of the indicators of financial sector development on economic growth. This might be because autocorrelation has not yet been tested. If on testing, autocorrelation exists, then it will have to be corrected.

5.1.C TESTING FOR AUTOCORRELATION

In case autocorrelation exists in a model, then it means that it is not in line with the assumption classical linear regression model due to the fact that errors are interdependent. Therefore, it becomes necessary to test for autocorrelation. One of the widely used test is the Durbin-Watson test. If this value is between 1 and 3, but nearer to 2, then it means that there is an acceptable level of autocorrelation and that it will not be a problem for the model. In our case, the DW-Statistic (Table 6.5) is 3.4261, thus slightly greater than 2. There is a sign of the presence of autocorrelation. Thus, the need to correct it arises.

5.1.D CORRECTING AUTOCORRELATION

From the test in 6.1.C, it is obvious that autocorrelation is present in the data. In order to correct this, the Cochrane – Orcut method will be used. This is an iterative procedure and it can be performed at either order 1, 2 or 3, whichever is making the data become more significant and the overall model more reliable. In this case, we will use AR(3). This process will be done on the Microfit Software itself.

Thus, the equation is still:

DRGDP = α + β₁DRM2 + β₂DRLIQUID + β₃DRMKTCAP + β₄DRGDFCF

When running this regression under the Cochrane – Orcut process, the results obtained are as follows (original results in Appendix):

Table 5.6: Regression results under Cochrane Orcut Process using AR(3)

REGRESSORS	COEFFICIENT	PROBABILITY
Α	0.030672	0.013
DRM2	-1.1489	0.018
DRLIQUID	-0.79927	0.004
DRMKTCAP	-0.072079	0.338
DRGDFCF	2.3177	0.000

R²	0.74456
DW-Statistic	2.3824

Source: Computed

Dependent variable is DRGDP

19 observations from 1987 to 2005

From the results above, it can be seen that after correcting for autocorrelation, most of the variables become significant except for DRMKTCAP which has p-value at 0.338, thus greater than 0.05. Also, the model becomes more reliable as R² is approximately 74%. Finally, the DW statistic also converges to 2, thus showing that autocorrelation has been reduced.

5.1.E DROPPING OF INSIGNIFICANT VARIABLE

Since the variable DRMKTCAP is not significant, we will drop it and see the effect on the regression. The equation is:

DRGDP = α + β₁DRM2 + β₂DRLIQUID + β₃DRGDFCF

The regression results are as follows:

Table 5.7: Regression results under Cochrane Orcut Process using AR(3)

(After dropping out DRMKTCAP)

REGRESSORS	COEFFICIENT	PROBABILITY
α	0.02899	0.018
DRM2	-1.1493	0.020
DRLIQUID	-0.78197	0.005
DRGDFCF	2.2618	0.000

R²	0.74581
DW-Statistic	2.1568

Source: Computed

Dependent variable is DRGDP

5.1.F ANALYSIS

From the results above, our final model can now be written as:

DRGDP = α + β₁DRM2 + β₂DRLIQUID + β₃DRGDFCF

It must be noted that this equation is obtained by checking for multicollinearity (RM2 and RLIQUID were highly correlated but this is not the case for DRM2 and DRLIQUID, that is why none was removed from the model at first), removing problems of non-stationarity, autocorrelation and removing any insignificant variable.

The equation then becomes (Table 6.7):

DRGDP = 0.02899 - 1.1493DRM2 - 0.78197DRLIQUID + 2.2618DRGDFCF

First of all, we will analyse the p-value of the variables (Table 6.7). The probability should be less than 0.05 in order that the variables are significant. In fact, under this regression, all variables are statistically significant. It must be noted that DRLIQUID and DRGDFCF are significant at 1%. This way, the analysis will be able to be carried out under this regression.

The constant in the model, that is α is positive at 0.0289 and implies that there are other variables which affect economic growth in Mauritius. In case the difference in these other variables was increased by 1 unit, this would lead to an increase in DRGDP, thus the difference in average growth by approximately 0.028, ceteris paribus.

As for DRM2, it is significant at 0.02, thus at less than 0.05. If the difference in real money stock ratio rises by 1 unit, then DRGDP will experience a fall by 1.15 approximately, thus a more than proportionate fall in DRGDP. As we can see, the predicted sign for DRM2 was positive but in fact the regression shows that there has been a negative effect on economic growth in Mauritius.

In the case of DRLIQUID, it is significant as the probability is less than 0.05 (0.005). However, in this case too, it was predicted that banks and other financial liabilities in an economy would be profitable and thus have a positive effect on economic growth. But the sign for DRLIQUID is negative. In case the difference in the real ratio of liquid liabilities rises by 1 unit, DRGDP will fall by 0.78.

DRGDFCF is the variable which has been significant in most regressions. The value is almost zero and this means that it is appropriate for this study. Not only that, the predicted sign was positive and in fact, the actual sign too is positive thus showing that when the difference in the real ratio of GDFCF rises by 1 unit, then DRGDP rises more than proportionately by 2.26 approximately.

It is important to note that when autocorrelation has been corrected, the DW statistic fell from 3.4 to 2.1568, thus converging more to the value of 2, indicating that there is no problem of the correlation of residuals.

Also, the adjusted coefficient of determination, also known as the R-Bar-Squared is at 74% approximately. As it gets closer to 100%, it means that the model is 100% right. This shows that this model is reliable at 74% and it explains any variation in the dependent variables. It therefore suggests that the sample regression line fits the data it can be concluded that the variables in fact affect the dependent variable.

5.2 MODEL 2

As we have seen, Model 1 is free from nonstationarity, multicollinearity and autocorrelation problems. Departing from this model, we now remove 1 variable and run the regression again in order to see whether better results can be obtained. In this case, we will remove DRM2. This model too will not have stationarity and multicollinearity problem. In order to remove autocorrelation, the regression will be run under the Cochrane Orcut Process at order 3. Thus the equation will be as follows:

DRGDP = α + β₁DRLIQUID + β₂DRGDFCF

When this equation is run under the Cochrane Orcut Procedure, the results obtained are:

Table 5.8: Regression results under Cochrane Orcut Process using AR(3)

(After excluding DRM2)

REGRESSORS	COEFFICIENT	PROBABILITY
α	0.0010	0.866
DRLIQUID	-0.51020	0.153
DRGDFCF	1.3436	0.008

R²	0.6359
DW-Statistic	1.9255

Source: Computed

Dependent variable is DRGDP

The equation becomes:

DRGDP = 0.0010 - 0.51020DRLIQUID + 1.3436DRGDFCF

When excluding the variable DRM2, only the variable DRGDFCF is significant. The constant and DRLIQUID are not significant as they are less than 0.05. The same sign result is obtained like Model 1. In this case, if DRLIQUID rises by 1 unit, DRGDP falls by 0.5 and if DGDFCF rises by a unit, then DRGDP rises by 1.34. also this model is reliable at 64% approximately.

5.3 MODEL 3

The next model which will be be constructed will be by removing the variable DRLIQUID and then by running the regression. The equation will be:

DRGDP = α + β₁DRM2 + β₂DRGDFCF

The regression will be done once again under the Cochrane Orcut Process to remove autocorrelation. The results are as follows:

Table 5.9: Regression results under Cochrane Orcut Process using AR(3)

(After excluding DRLIQUID)

REGRESSORS	COEFFICIENT	PROBABILITY
α	0.021267	0.281
DRM2	-1.3262	0.086
DRGDFCF	2.0430	0.002

R²	0.66444
DW-Statistic	2.1057

Source: Computed

Dependent variable is DRGDP

The equation becomes:

DRGDP = 0.02126 - 1.3262DRLIQUID + 2.0430DRGDFCF

When removing DRLIQUID, the regression results show that DRM2 is not significant. However if compared at 0.1, it is significant. DRGDFCF is also significant at 5%. With a rise in 1 unit in the first difference of real ratio on money stock to GDP, DRGDP falls by 1.32. .As for DRGDFCF, if it increases by 1 unit, DRGDP rises by 2.04.

This model has no autocorrelation as DW-Statistic is near to 2. Also this model is reliable at 66%.

5.4 MODEL 4

Model 4 will be estimated by dropping out DRGDFCF and by following the same process as Model 2 & 3, i.e by running the regression under the Cochrane Orcut Process. The equation is thus:

DRGDP = α + β₁DRM2 + β₂DRLIQUID

The regression results under the Cochrane Orcut Process are shown below:

Table 5.10: Regression results under Cochrane Orcut Process using AR(3)

(After excluding DRGDFCF)

REGRESSORS	COEFFICIENT	PROBABILITY
α	-0.02553	0.170
DRM2	0.79322	0.259
DRLIQUID	0.40847	0.386

R²	0.41252
DW-Statistic	2.0940

Source: Computed

Dependent variable is DRGDP

The regression results are as follows:

DRGDP = - 0.02553 + 0.79322DRM2 + 0.40847DRLIQUID

When excluding DRGDFCF, all variables become insignificant. However, both variables DRM2 and DRLIQUID are positive, just like the predicted sign value. Also, removing the ratio of investment to GDP reduces the reliability of the model at 41%. The DW Statistic is at 2.094, thus showing no autocorrelation.

5.5 LIMITATIONS

As we have seen, results from the various regressions depend largely on the data that has been collected, the methodology used and the tests undertaken. However, it must be noted that these regressions consisted of time series data of a range of only 20 years. In case more observations were used, the results might have been different and more reliable.

Furthermore, only 6 variables have been used and while testing for multicollinearity, autocorrelation and nonstationarity, many had to be dropped, thus having a model with only 3 variables at last. The results would have been more conclusive in case more variables had been used as economic growth depends on many other factors.

CHAPTER 6: CONCLUSION

From the four models designed in the previous chapter, Model 1 is better as it includes more variables, all of which are significant and it has higher explanatory power.

Also, from the regressions performed above, it has been noticed that DRGDFCF has got the maximum explanatory power. It has been significant in most regressions, and it has obtained the predicted positive sign. It has been the only variable to affect DRGDP positively. Thus government policies should be aimed towards increasing investment in the country. Rate of interest in the country should be increased in order to allow people to save money. This savings will in turn become investment in the country and finally it will have an impact on the GDP.

As for DRLIQUID and DRM2, both have been significant but there have been deviations concerning the sign of the coefficients. Both have been seen to have a negative impact on DRGDP. This is a cause of concern for the country as the financial sector of Mauritius is largely dependant on the banking sector.

However, it must be noted that there are many other factors which affect the financial sector which in turn has an impact on economic growth. The global business sector in Mauritius is facing tremendous makeover and expanding really rapidly. This is a good sign as this shows that Mauritius is moving towards a more liberalised financial sector rather than depending on the usual financial intermediaries.

Therefore, it becomes vital that the government encourages a closer integration of the onshore and the offshore activities. Steps have already been taken as the financial sector has recently experienced reforms in the legislations. New laws have been passed in order to increase the competitive global environment in which the financial sector finds itself today.

Custom Oxbridge Essays & Oxbridge Dissertations by Oxbridge Graduates.

Financial Growth Mauritius

Undergraduate Services

Masters Services

PhD Services

Law Services

Quick Links

Study Aids

Free Essays

Oxbridge Writers: The UK's number one provider of custom essays from Oxford and Cambridge University qualified experts - Oxbridge Essays Service