The Impact of Monetary Policy Announcements by the South African Reserve Bank on Stock Market Returns Using Forward Rate Agreements Submitted by: Nhlamulo Collins Mabasa Student no: 692475 A dissertation submitted to the Wits Business School, University of the Witwatersrand, Johannesburg, South Africa, in fulfilment of the requirements for the Master of Management in Finance & Investment. Supervisor: Dr. Euphemia Godspower-Akpomiemie Johannesburg: March 2021 i DECLARATION Declaration 1. I know that plagiarism is wrong. Plagiarism is to use another’s work and pretend that it is one’s own. 2. I have used a recognized convention for citation and referencing. Each contribution to, and quotation from the work(s) of other people has been attributed, and has been cited and referenced. 3. This essay/report/project is my own work. 4. I have not allowed, and will not allow, anyone to copy my work with the intention of passing it off as his or her own work. 5. I acknowledge that copying someone else’s assignment or essay, or part of it, is wrong, and declare that this is my own work. Signature ______ ________________________ Nhlamulo Collins Mabasa ii ABSTRACT The objective of this paper is to explore the unanticipated impact of monetary policy announcements on stock market returns using Forward Rate Agreements (FRAs). This paper looks at the Johannesburg Stock Exchange (JSE) All Share Index and two other sectoral specific stock market indices (Financial and Industrial sector indices) and assess the responsiveness of these stock market returns to unanticipated monetary policy announcement shock. In an attempt to understand this relationship between monetary policy and the stock market, the main empirical view suggests that decomposing monetary policy changes into anticipated and unanticipated components is crucial for discerning their effects. The decomposition of unexpected policy rates is based on the futures market. In the absence of the South African interest rate futures market, this study employs a FRA which serves as a measure of monetary policy surprise. This study begins with a 1-day event study, which examines the immediate impact of monetary policy shocks on the stock market, and then use an Ordinary Least Squares (OLS) regression analysis, which provides insight to into the dynamic effects of the unanticipated interest rate shock on the stock market. This study employs a time series percentage change daily closing prices of the South African stock market (JSE all share Index, sub-indices), actual changes in the South African Reserve Bank(SARB) repo rate, and the FRA, spanning the period January 2010 through to December 2019, which explores the post-global recession dynamics. The study shows that a hypothetical unanticipated increase of 1% repo rate results in a decline of 0.32 percentage of the Johannesburg Stock Exchange ALL SHARE Index. The findings and recommendations are crucial for the South African central bank authorities and stock market participants as it explains the process through which monetary policy outcomes are transmitted to the real economy, inflation and employment. A future piece of work could contemporarily assess the impact of monetary policy and other sector related and political events on the stock market iii DEDICATION First and foremost, I want to thank God for empowering me with knowledge to withstand this work during uncertain and abstract times. I dedicate this Dissertation to my mother for her love and support throughout my life. Thank you for giving me strength to reach for the stars and chase my dreams. iv ACKNOWLEDGEMENTS I would like to thank my supervisor Dr. Euphemia Godspower-Akpomiemie for her great guidance and support during this study, without her this would not have been possible. I would also like to thank my sisters (Khanyisa and Yvonne), who without this would have not been possible. My sincere thanks go to Mr Mothupi Modiba for having shown faith in me by pledging financial resources for me to pursue this Master’s degree, without his generous hand, I wouldn’t be a Masters holder today. Lastly, I want to thank my good friends, Russel Mulamula, Vukosi Era Maluleke, Nathi Ndlovu and Vukosi Sambo, thank you for taking interest in my work. v Table of Content DECLARATION .................................................................................................................................... i ABSTRACT .......................................................................................................................................... ii DEDICATION ..................................................................................................................................... iii ACKNOWLEDGEMENTS ..................................................................................................................... iv LIST OF FIGURES ............................................................................................................................... vii LIST OF TABLES ............................................................................................................................... viii LIST OF SYMBOLS .............................................................................................................................. ix NOMENCLATURE ............................................................................................................................... x Chapter 1: Introduction ......................................................................................................................1 1.1 Context of the study .................................................................................................................1 1.2 Problem Statement ...................................................................................................................4 1.3 Research Objectives ..................................................................................................................5 1.4 Significance of the study ...........................................................................................................6 1.5 Outline of the Study ..................................................................................................................6 Chapter 2: Brief Literature Review ......................................................................................................7 2.1 Introduction .............................................................................................................................7 2.2 Efficient Market Hypothesis consistent ......................................................................................7 2.3 Quantity Theory of Money ...................................................................................................... 10 2. 4 Standard Valuation Model: Discounted Cash Flow Model ........................................................... 12 2.5 Central Bank Communication and Announcement effect ......................................................... 13 2.6 Monetary Policy Transmission Channel to the Stock Market and the Real Economy ................. 15 2.6.1 Empirical evidence from the Developed Economies ........................................................... 18 2.6.2 Empirical evidence from the Developing Economies .......................................................... 20 2. 7 Application of Forward Rate Agreements ................................................................................... 22 2.8 Overview of the South African Monetary Policy in last 10 years ............................................... 24 2.9 Chapter Summary ................................................................................................................... 30 Chapter 3. Data Description and Research Methodology ........................................................... 31 3.1 Introduction ........................................................................................................................... 31 3.2 Research Design ..................................................................................................................... 31 3.3 Population and Sample ........................................................................................................... 31 vi 3.4 Data Collection ....................................................................................................................... 34 3.5 Measurement of Variables ...................................................................................................... 34 3.6 Data Analysis .......................................................................................................................... 35 3.6.1 Event Studies ................................................................................................................... 35 3.6.2 OLS Regression ................................................................................................................. 37 3.7 Ethical Considerations ............................................................................................................. 39 3.8 Validity and Reliability ............................................................................................................ 39 3.9 Chapter Summary ................................................................................................................... 40 Chapter 4: Presentation of Results .................................................................................................... 41 4.1 Introduction ........................................................................................................................... 41 4.2 Descriptive statistics ............................................................................................................... 41 4.3 Pre Estimation Analysis ........................................................................................................... 42 4.4 Baseline specification Results .................................................................................................. 46 4.5 Summary ................................................................................................................................ 49 Chapter 5: Summary and Conclusions ............................................................................................... 51 5.1 Introduction ........................................................................................................................... 51 5.2 Discussion of the objectives & Conclusion ............................................................................... 51 5.3 Policy Recommendations ........................................................................................................ 53 5.4 Recommendation for further research .................................................................................... 54 References: ...................................................................................................................................... 55 Appendix ......................................................................................................................................... 63 vii LIST OF FIGURES Figure 2.1 Monetary Policy Transmission Mechanism…………………… Page 22 Figure 2.2 SARB Repo Rate…………………………………………………… Page 33 Figure 2.3 Inflation and GDP Growth…………………………………… Page 34 Figure 2.4 Unemployment Rate……………………………………………… Page 35 Figure 3.1 Stock Market Variables…………………………………………… Page 38 Figure 3.2 Repo Rate and FRA2x5………………………………………… Page 40 Figure 3.3 Event Window…………………………………………………… Page 42 Figure 4.1 JSE All Share and Repo Rate………………………………… Page 48 Figure 4.2 JSE All Share and FRA 2x5…………………………………… Page 49 Figure 4.3 Residuals Plot …………………………………………………… Page 50 Figure 4.4 Normality Plot of Residuals-Histogram………………………… Page 51 viii LIST OF TABLES Table 2.1 MPC Scheduled and Unscheduled Meetings…………………… Page 32 Table 4.1 Descriptive Statistics………………………………………… Page 47 Table 4.2 Variance Inflector Factor……………………………………… Page 49 Table 4.3 Unit Root Test Results………………………………………… Page 51 Table 4.4 OLS Regression Results: JSE ALL Share Index….……………… Page 52 Table 4.5 OLS Regression Results: JSE Financial Index…………………… Page 54 Table 4.6 OLS Regression Results: JSE Industrial Index…………….…… Page 55 Table A1: Summary of Estimates Results………………………....................... Page 69 Table A2: Test of Serial Correlation…………………………………………... Page 70 Table A3: MPC Announcement Dates……………………………………… Page 71 ix LIST OF SYMBOLS Δ Percentage Change x NOMENCLATURE ADF Augmented Dickey Fuller ANC African National Congress BA Bankers’ Acceptance BRICS Brazil, Russia, India, China, South Africa BVAR Bayesian Vector Autoregression CPI Consumer Price Index DW Durbin-Watson ECB European Central Bank EMH Efficient Market Hypothesis FED Federal Reserve FOMC Federal Open Market Committee FRA Forward Rate Agreement FEVD Forecast Error Variance Decomposition GARCH Generalized Autoregressive Conditional Heteroskedasticity GDP Gross Domestic Product JIBAR Johannesburg Interbank Agreement Rate JSE Johannesburg Stock Exchange MPC Monetary Policy Committee NDP National Development Plan OECD Organization for economic operation and development OLS Ordinary Least Squares REPO Repurchase Agreement S.A South Africa SARB South African Reserve Bank SVAR Structural Vector Autoregressive SVEC Structural Vector Error Correction UK United Kingdom USA United States of America VAR Vector Auto Regressive 1 Chapter 1: Introduction 1.1 Context of the study The majority of countries in Africa have encountered many economic challenges over the years, and regular changes to monetary policy have not yielded any desired economic results (Iddrisu, Harvey & Amidu, 2016). This raises many questions about the effectiveness of monetary policy in stimulating economic activities. Aziza (2010) defines monetary policy as a tool utilised by Central banks to influence the required rate of inflation, growth in real economic activity, stabilisation of the exchange rate, real output, employment. As a result, monetary policy outcomes tend to affect the broader financial markets ranging from the stock market, foreign exchange market, and the bond market to list a few. The overriding objective of monetary policy is usually to maintain and restore financial market stability within the economy (Bissoon et al., 2016). Such policies can either be expansive or restrictive depending on the type of instruments (interest rates and money supply) used by central banks (Bissoon et al.,2016). Expansionary policy is achieved through decreasing interest rates or increasing the stock of money in circulation in the economy. This is done traditionally to stimulate the economy and also used to increase employment opportunities in a recession (Aziza, 2010). Whereas, restrictive policy on the other hand is achieved by increasing interest rates or decreasing the money in circulation in the economy. This is done to keep inflation within the targeted range. Monetary policy is implemented by the central bank to realise a set of objectives that are demonstrated in terms of macroeconomic variables such as inflation, real output, and employment (Ioannidis & Kontonikas, 2008). However, monetary policy instruments are found to influence these variables at best indirectly (Bernanke & Gertler, 2005). Petelis (1997) shows that monetary policy decisions play a significant role in predicting asset returns in the United States of America (USA). He contends that a relationship between future expected excess stock returns and monetary policy variables exists. The most immediate and direct impact of monetary policy actions is on the financial markets through changes in the federal funds rate which affects asset returns and prices (Bernanke & Gertler, 2005). 2 Monetary policy tools used by the central bank include interest rates, reserve requirements (cash requirements or cash ratio and liquidity), rediscount rate, selective credit controls, treasury rate amongst others (Aziza, 2010). Stock prices can be affected by changes in interest rates in two ways: firstly, it affects the discount rate which financial market participants use to calculate the present value of the firm’s future cash flows, and secondly, it influences expectations of future performance of companies (Schrey & Hafdísarson, 2017). Bernake & Gertler (2005) holds a view that, in the context of short-term monetary policy management, central banks should pursue a unified policy framework that treats financial stability and price stability as highly complementary and mutually consistent. The global equity market volatility that was experienced in 2007/08 and the economic instability that followed revived the debate on whether stock prices influence economic activities as well as the role of monetary policy in protecting economies against disruptive effects of stock market volatility (Bonga-Bonga, 2012). Apart from the 2007/08 financial crisis that befell global economies, previous centuries recorded two other major economic events of unexpected asset price reversals after sustaining long periods of stock price increases. The two major events were; 1929 US market crush and the Japanese experience of the late 1980s and early 1990s. Both events were characterised by asset price boom- bust cycles, which is the decline in asset prices that triggered increased banking and financial sector instability that consequently slowed down real economic activity (Ioannidis & Kontonikas, 2008). Bernanke & Gertler (2005) points out that the volatility in asset prices has been a major preoccupation for academics as it has the potential to destabilise financial markets. Although there is wide consensus on what constitutes volatility in the stock market, to a lesser degree to quantify it, the causes of changes in stock market volatility have generated huge debates (Mala & Reddy, 2007). Some argues that the changes in stock price returns is as a result of unanticipated information or events like the terrorist attack of 2001 in the United States, while some claim that volatility could be due to changes in trading volumes, patterns, or practices (Mala & Reddy, 2007). 3 However, it is strongly argued by most economists that monetary policy has a strong influence on stabilising financial market activities. Under new Keynesian theory, central banks through monetary policy tools can exert some level of control on real interest rates as a result of sticky prices in the short run (Bjørnland & Leitemo, 2009). In the emerging market context, the role of the stock market in influencing real economic activity is increasing, hence proponents are arguing for a much proactive role of central banks (Muroyiwa, Ezeoha, & Mushunje, 2017). From the viewpoint of monetary policy setters, asset prices are a key determinant in the evaluation of the transmission process of interest rates (Bohl, Siklos & Sondermann, 2008). Hojat (2015) argues that changes in monetary policy both quantitative (through changes in interest rates) and qualitative methods (through wording in statements released by central banks) provides forward guidance about the future economic path. Information communicated by central banks about monetary policy decisions is incorporated by financial markets to improve investment decisions, which eventually reflect stock prices (Gupta & Reid, 2013). Central banks have moved towards greater transparency in the past decade by increasing the channels of communication and the level of information provided to the public (Amato, Morris, & Shin 2002). The effectiveness of monetary policy is not only measured by the instruments they employ to execute their mandate, but also involves a high degree of transparency in communicating their decisions. Blinder et al., (2008) highlight that, independent central banks have a responsibility to explain their decisions and rationale underpinning their actions to increase transparency in their operations. Before period 1994, central banks across the globe were not transparent in the way they used to implement monetary policy. It was only after February 1994, when the Federal Open Market Committee (FOMC) began publishing in its statements an assessment of its bias concerning future changes in monetary policy (Blinder et al., 2008). Central bank communication has evolved as a significant instrument for central bankers over the past 15 years (Neuenkirch, 2013). Blinder et al., (2008) define Central bank communication as the provision of information to the public highlighting the monetary policy strategy, objectives, economic outlook, and the outlook for future 4 policy actions. By providing information to the public, Central banks can influence economic activity as market participants continuously incorporate information provided in their investment decisions. Lately, it is broadly accepted that central banks can affect the economy by influencing market expectations about the future path of overnight interest rates (Blinder et al., 2008). This research provides an empirical analysis on the effect of interest rate announcement on the South African (S.A) stock market between 2010 and 2019. It establishes a link between monetary policy outcomes and financial assets which is determining the role of the transmission mechanism in economic development (Ioannidis & Kontonikas, 2008). 1.2 Problem Statement The South African Stock Exchange, popularity known as the JSE is by far the largest and most liquid stock market in Africa which makes it the most lucrative investment destination of market investors in the equity space (Muroyiwa, Ezeoha, & Mushunje, 2017). By the end of October 2018, the JSE market capitalization stood at around 891.71 USD Billion which records a decrease of from the previous number of 988.339 USD Billion for September 2018 (CEICDATA, 2018). The JSE has consistently been in the top 20 most capitalized stock exchange and also falls within the top 6 of emerging market economies with more than 400 firms listed (Hassan, 2018). The South African stock market importance to the national economy, measured by the ration of the market capitalization to the GDP is at 332% in 2018 and this is unusually large which placed it second largest in the world (Bloomberg, 2018). The mining sector is both the foundation of the bourse and a reason for development and growth of the South African financial services sector (Hassan, 2013). The role played by equity market in South Africa is also increasing by day, hence it is important for the SARB to understand the role and impact the JSE has on their decisions, the economy and ultimately goals they seek to achieve. A considerable amount of theoretical and empirical contributions suggest that stock prices play a crucial role in economic development via wealth and balance sheet effects (Kontonikas & Montagnoli, 2008; Ioannidis & Kontonikas, 2008; Bjørnland & Leitemo 2009). Changes in monetary policy outcomes affect investor’s decisions and stock market returns. Based on the 5 literature review, studies carried out in other countries across different economies have empirically displayed that macroeconomic policy outcomes influence the performance of the stock market. Studies by Kuttner (2001); Bernanke & Kuttner (2005); Bredin, Gavin & O Reilly (2003); Wiranto (2008), Rigobon & Sack (2002) did event studies that discriminated between anticipated and unanticipated changes in the policy instrument to assess the impact of monetary policy on the stock market. It is argued that studies that fail to isolate monetary changes into the anticipated and unanticipated components are most likely to record biased results due to an error in variables problems (Bredin, Gavin & O Reilly, 2003). In the South African context, a recent study has investigated the sensitivity of industry stock returns to monetary policy and macroeconomic news (Gupta & Reid, 2013). Their study was able to decompose the expected and unexpected components of monetary policy surprise by using the change in the 3 months Banker’s acceptance (BA) rate on the day of the day of MPC meeting announcement. Another study by Ramatlo (2019) was able to make a distinction between anticipated and unanticipated policy changes and found that a surprise 100 basis points increase cause a short-term decrease of the JSE ALL Share index by 2.71%. Her study used Johannesburg Interbank Agreed Rate (JIBAR) to capture the unanticipated component of monetary policy. To the best of my knowledge, there is no study that has examined the impact of monetary policy announcements on the South African stock market performance using Forward Rate Agreements (FRAs) to identify unanticipated monetary policy announcement shocks post the global financial crisis of 2007/08. However, a related study to this topic was done by Aron & Muellbauer (2007) who used FRAs as a proxy for expectations on future interest rates to assess the credibility and predictability of monetary policy since the adoption of inflation targeting framework in S.A. Therefore, this study was considered to empirically identify, investigate, and understand the impact of monetary policy decisions on stock market returns in South Africa using the FRAs. 1.3 Research Objectives Based on the identified problems, the objective of this study is to assess the impact of monetary policy announcements on the South African stock market at an aggregate and industry specific level in an event type study between period 2010 and 2019. Firstly, the study decomposes policy 6 rate changes for South African stock market returns into their expected and unexpected components on the day of the policy change on information gathered from forward rate agreements. Next, I regress these surprise and anticipated repo rate changes on the Stock market returns on the day of the change in the respective policy rate cycle. Ultimately, this would aid to material stakeholders within the market environment to make investment decisions based on a balance of risks decision attached to the stock market reaction after the monetary policy announcement has been made on the last day of the Monetary Policy Committee meeting. 1.4 Significance of the study This study analyses the impact of monetary policy announcements on stock market returns in South Africa. It deepens our knowledge of the nature of transmission mechanism of monetary policy in the South African economy. This study also provides additional information on financial asset investors (Equity and Bonds Investors), policymakers, the academic community, and other stock market stakeholders on how monetary policy announcements by the SARB impact stock market returns. Moreover, it further provides investors with information on how monetary policy outcomes alter the performance of their portfolios, therefore will aid investment decisions. 1.5 Outline of the Study This study was organised into five chapters. The first chapter constituted an introduction to the study which comprised the background to the study, problem statement, research objectives, and significance of the study. The second chapter presented the theoretical framework for the monetary policy transmission mechanism, review of literature on various papers, journals, theses that discuss monetary policy outcomes and the stock market. The third chapter discussed the methodology and the estimation process for the impact of monetary policy outcomes on stock market returns in South Africa. The fourth Chapter covered the results and the discussions on the findings of the study. Chapter Five focused on the research summary, findings, conclusion, and recommendations. 7 Chapter 2: Brief Literature Review 2.1 Introduction The relationship between monetary policy decisions and financial asset prices has been studied in academia at great length across the world. The causal complexities between monetary policy outcomes and asset returns have become an interesting subject matter for financial economists as policy changes affect investor’s required rate of return on assets (Stoica & Diaconasu, 2012). The role of monetary policy in determining equity returns is either through changing the discount rate or by influencing the expectations of market participants about the future of economic activities (Ioannidis & Kontonikas, 2008). Having discovered contradictions in the literature concerning the relationship between monetary policy outcomes and the stock market performance, this section discusses different academic positions to critically evaluate the arguments and findings pronounced in different studies and economies the world over. 2.2 Efficient Market Hypothesis consistent During the 1960s, the efficient market hypothesis (EMH) was introduced into the literature of financial economics and the theory was coined to refute claims of technical stock analysis which predicates future price movements on past price patterns of stocks (Schrey & Hafdísarson, 2017). Fama (1995, p.76) defines an “efficient market hypothesis as a market where there are a large number of rational Profit-Maximiser’s actively competing, with each other trying to predict future market values of individual securities, and where important current information is almost freely available to all participants”. It is generally believed when financial market participants become aware of new information, the news will quickly be reflected into the prices of securities without any delay and consequently, stock prices should be consistent with fundamentals. The EMH in the modern stock market is described by trading conditions due to the free flow of information and trade execution is faster than ever (Degutis & Novickytė, 2014). 8 Apart from the definition of efficient markets, Fana (1970) produced an article that provided a distinction between three forms of efficiency- weak, semi-strong, and strong form. The weak form efficient market hypothesis is consistently tied to the idea of a “random walk. In the finance literature, random walk is described as a price series where all successive asset price variations represent arbitrary departures from previous prices (Malkiel, 1989). The rationality around the random walk theory is rooted in the idea of a free flow of information and the immediate reflection of such information on stock prices once available to market participants. Therefore, it further implies that tomorrow’s changes in asset prices reflect tomorrow’s information independent of the changes in asset prices of today. Bodie, Kane, and Marcus (2014) gave a detailed description of market efficiency as follows: The weak-form hypotheses state that market trading data such as the historical prices, short interest, and trading volume will be reflected in today’s stock market levels. This form of market efficiency denotes that trend analysis is regarded as ineffectual. Previous stock price data is freely available to the public and at no cost, and therefore this hypothesis indicates that all investors already would have learned to exploit the signals derived from the publicly available data to predict future stock market performance. Eventually, the signals become obsolete as they are broadly identified because a sell signal, for example, would result in an instant price decrease. The semi-strong form hypotheses assert that stock prices must reflect all publicly available information regarding the prospects of a firm. The said information includes amongst others quality of management, balance sheet composition, patents held, earnings forecast, accounting practices, product line. Again, if financial market participants freely have access to this kind of information from publicly available sources, it should be expected that it will be reflected in stock prices. Lastly, the strong-form hypotheses highlight that not only reflects all information relevant to the firm but also information available only to insiders privy to the inner workings of a firm. This market hypothesis is rather extreme and it could be argued that insiders are fortunate enough to gain access to pertinent information in advance before public release could use that advantage to profit from trading on that information. 9 These theories demonstrate that the financial markets cannot function optimally without data and therefore the distribution of information is illustrious through different forms of market efficiencies. Investors rely on information to optimally allocate savings by accurately estimating the future values of a financial asset. The efficient market hypothesis was tested in many capital markets across global economies, thus demonstrating different outcomes. Examining the Istanbul Stock Exchange (ISE) on a monthly return index-20 from January 1986 to November 2005, Aga & Kocaman (2008) concluded that the time series analysis demonstrates that the returns can be described only by the constant term, which is mean and there is a weak form of efficiency in ISE. Meaning that the market is weakly efficient if the current prices cannot be explained with the historical values. Findings by Borges (2010) on the efficient market hypothesis for six European stock markets (Greece, France, Germany, UK, Portugal, Spain) from January 1993 to December 2007 showed evidence that monthly prices and returns follow random walks in all six countries. However, Daily returns demonstrate mixed results. Daily returns are not normally distributed due to being negatively skewed and leptokurtic; Germany, UK, and Spain display a random walk behavior whilst the hypothesis did not meet most of the criteria for Portugal and Greece as a result of serial positive correlation. Guduza & Phiri (2017) investigated a weak form efficiency for 4 stock and 7 bond market return on the JSE from 2002 to 2016. The study used both individual and panel-based unit root testing methods to overwhelmingly conclude that there is evidence of weak-form efficiency as the integration test failed to establish evidence of unit root behavior amongst the time series data observed. Therefore, the study by Guduza & Phiri (2017) on the efficiency of debt and stock markets in the South African economy given the global financial crisis is thus confirmed. A study done on the South African stock market by Van Heerden et al., (2013) examined the efficient market hypothesis on the JSE between 2001 and 2013. It is worth noting that the study deviated away from the orthodox use of the linear approach to employ the Threshold Autoregressive (TAR) model and corresponding asymmetric unit root test to show how the stock market progresses as highly persistent and nonlinear. The study found that for most of the time series under observation, 10 the stationarity hypothesis among the variables is rejected by the formal unit root test. These findings bridge two contrasting views derived from previous studies by deducing that even though several stock prices on the JSE may not change as pure unit root methods, the time series are, conversely, highly robust to be regarded as weak-form efficient. 2.3 Quantity Theory of Money There are at least two consistent explanations regarding the assumptions that stock prices are negatively related to unanticipated announced changes in money supply (Pearce & Roley ,1983; Lucas, 1980). The first explanation is that the unexpected rise in money supply, market participants revise upward their future inflation expectations. Consequently, there are several channels in which the rise in inflation expectation subdues the equity returns. Feldstein (1983), argues that the reason why the expected rise inflation depresses the stock market is due to the fact that inflation raises effective tax rate on corporate source income. Lower expected corporates profits require stock prices to fall in order for stock to regain its competitive advantage. The second channel relates to the notion that the rise in inflation expectation would cause the equity prices to fall as it raises the expected returns on alternative assets such as real estate ((Pearce & Roley ,1983). Lastly, Feldstein (1993) contends that equity prices are subdued because the rise in expected inflation raises the rate of interest rate that can be derived by investing in nominal return on bonds. He argues that this argument be overruled since the higher nominal rate of interest rate generally corresponds to unchanged real rate of interest rate. According to Pearce & Roley (1983), the second explanation is that the response of equity prices to unexpected money supply reflects market participant’s expectations of the reaction of the Federal Reserve Bank to the surprise. Lucas (1980, p.1005) posits that these assumptions outlines above “possess a combination of theoretical coherence and empirical verification shared by no other propositions in monetary economics”. In Particular, central banks increases short term interest rates to offset the surge in money. Furthermore, with lagged reserve accounting, short term interest rates may rise even in the absence of central banks actions if market players increase their assessment of the excess demand for reserves (Pearce & Roley, 1983). 11 The first step in assessing the reaction of the stock prices to announced changes in money is usual linear model is employed as represented by Pearce & Roley (1983, p. 4) using the following models: ∆𝑆𝑃𝑡 = 𝑎 + 𝑏 (∆𝑀𝑡 𝑎 -∆𝑀𝑡 𝑒) +𝑒𝑡 Where ∆𝑆𝑃𝑡 = Change in stock prices observed after the money announcement ∆𝑀𝑡 𝑎 = announced change in the money stock ∆𝑀𝑡 𝑒= expected change in the money stock 𝑒𝑡 = random error term The basic proposition of the efficient markets theory is that only the unexpected change in money should influence stock prices so that the data should not reject the restriction that the coefficients on ∆𝑀𝑡 𝑎 and ∆𝑀𝑡 𝑒 sum to zero. The hypothesized behavior of security market participants outlined above further stipulates that b should be negative However, assertions about the theory of money assumptions discussed above by Pearce & Roley (1983) have been empirically and theoretically critisised by other economists. Firstly, Smith, (1988) argues that in order for the theory of money propositions to hold, monetary changes need to have a material impact on the consolidated balance sheet of the Nations treasury and central bank. However, monetary actions that fails to alter this consolidate balance sheets can be immaterial for stock prices and interest rates. Sargent & Smith (1988) demonstrated this point by providing a once and for all change in money stock that produced a no effect impact on prices or interest rates. One of their finding is that central banks open market operations accomplished with fiscal policy held constant (that occur with the consolidated balance sheet of the Central bank and National Treasury unchanged) have no impact on prices. Secondly, they found that Governments attempt to intervene in the foreign exchange market can be effective if can only be effective if complemented by fiscal policy actions that have redistributive outcomes. Smith (1988) explains this further by highlighting that when central banks engage in in open market operations, they achieve that by exchanging interest bearing liabilities like bonds and non-interest bearing liabilities like currency. 12 2. 4 Standard Valuation Model: Discounted Cash Flow Model The Discounted cash flow model provides critical insights on the effects of monetary policy actions on the stock market. Ioannidis & Kontonikas (2008, p.4) expanded on this theory by providing a more detailed explanation: According to this widely used model, the stock price (𝑆𝑡) is the Present value of expected future dividends (𝐷𝑡+𝑗), Under the assumption of constant discount rate ( R ), It can be shown that 𝑺𝒕= 𝑬𝒕 [∑ 𝑘 𝑗=1 ( 1 1+𝑅 ) 𝐽 𝐷𝑡+𝐽] +𝑬𝒕 [∑ 𝑘 𝑗=1 ( 1 1+𝑅 ) 𝐾 𝑆𝑡+𝑘] (1) Where, 𝐸𝑡 is the conditional expectations operator based on information available in the market participants at time t, R is the rate of return used by market participants to discount future dividends, and K is the investor’s time horizon (stock holding period). The standard transvesality condition implies that as the Horizon k, increases the second term in the right hand sight of Eq. 1 vanishes to zero (no rational stock price bubbles): lim 𝑛→∞ 𝑬𝒕 [( 1 1+𝑅 ) 𝐾 𝑆𝑡+𝑘]= 0 (2) Thus, we obtain the familiar version of the present value model: 𝑺𝒕= 𝑬𝒕 [∑ 𝑘 𝑗=1 ( 1 1+𝑅 ) 𝐽 𝐷𝑡+𝐽] (3) Eq. (3) indicates that a change in monetary policy can affect stock returns in a dual manner. First, there is a direct effect on stock returns by altering the discount rate used by market participants. Tighter monetary policy leads to an increase in the rate at which firms’ future cash flows are capitalised causing stock prices to decline. The underlying assumptions are that, first, the discount factors used by market participants are generally linked to market rates of interest and second, the central bank is able to influence market interest rates. Second, monetary policy changes exert an indirect effect on the firms’ stock value by altering expected future cash flows. Monetary policy easing is expected to increase the overall level of economic activity and the stock price responds in a positive manner (expecting higher cash flows in the future). Hence, this channel generally assumes the existence of a link between monetary policy and the aggregate real economy. As 13 Patelis (1997) argues, stocks are claims on future economic output, so if monetary policy has real economic effects then stock markets should be influenced by monetary conditions. The application of this theory is further expanded in section 2.5 below. 2.5 Central Bank Communication and Announcement effect Since financial market participants are forward-looking, central banks affect the economy as much by influencing expectations through any direct, mechanical effects of central bank trading in the market for overnight cash (Woodford, 2005). Shaping and managing market expectations is an important part of monetary policy and this process is only possible with an effective channel of communication between a central bank and financial markets (Amato, Morris, & Shin, 2002). Central banks manage expectations by reducing noise or creating news. News creation focuses on how the central bank’s announcement influence expectations and therefore drive movements in asset prices (Blinder et al., 2008). Noise reduction focuses on the ability of the central bank to increase the predictability of its actions to reduce volatility in the financial markets (Blinder et al., 2008). This indicates that, if financial markets have all the information, they can accurately assess the developments of the economy and thus reduce levels of uncertainty in the market (Lehtimäki & Palmu, 2019). Woodford (2005) argues that the central bank’s willingness to share with the public information and assumptions on future policy increases the predictability of monetary policy to achieve financial markets stability. With the help of new technologies, the amount of information central banks produces and communicate has been expanded in response to the increasing weight of transparency (Lehtimäki & Palmu, 2019). Evidence by Blinder et al., (2008) indicates that communication can be a significant component of the central bank’s toolkit as it can influence movements in the financial markets, assist central banks to achieve their macroeconomic objectives, and potentially improve the predictability of monetary policy decisions. The study from Neuenkirch (2013) supports the idea of a relationship between monetary policy change signals and asset prices. The study analysed the Swiss Economic Institute’s Monetary Policy Communicator to measure the future path of the European Central Bank’s monetary policy and found that communication influences prices and output. The study 14 further shows that communication partially crowds out the impact of the short-term interest rate as the latter’s influence is lower and its implementation lag increases compared to a benchmark model without central bank communication. Luangaram & Sethapramote (2016) did a broad assessment of communication on Thailand’s monetary policy effectiveness looking at the following aspects, i.e., predictability of short-run policy interest rate, monetary transmission mechanism, and the ability to anchoring long-run inflation expectations. The study augmented the communication measure with various Taylor-type rule specifications and determined that monetary policy statements assist in improving the predictability of short-run interest rates. The study further used structural vector autoregression to find that when communication is incorporated, the impulse responses of policy rate shock on inflation and output are robust, demonstrating the enhanced effectiveness of the transmission mechanism process. Lastly, increasing interest rates can anchor inflation in the short run, whereas monetary policy communication drives long-term inflation expectations. A study by Waud (1970) supports the existence of a relationship between signals of changes in monetary policy and stock prices. He assessed the public interpretation of federal discount rate changes experimenting with the announcement effect. According to Schrey & Hafdísarson (2017), Market participants will always have expectations about future cash flows and economic trends. Waud (1970) argues changes in the discount rate have a psychological impact on Market wide expectations about the future performance of the economy. His study concludes that there is a notable effect of the announcement effect on expectations associated with changes in the discount rate and that days preceding to the reduction of the discount rate, the seems to exists evidence of the anticipation of change. The conclusion that can be drawn from Waud’s findings is that impact associated with interest rate outcomes and monetary policy decisions comprise of information that provide clues about the future economic conditions. Bomfim (2003) analyses pre-announcement and news effects on the stock market returns in the context of public disclosure of monetary policy decisions. His study focused on two vantage points: days around FOMC scheduled meetings and days of actual policy announcements involving the target level of the federal funds rate. He concludes that the equity market tends to be 15 moderately muted and conditional volatility is abnormally low on days prior to regular monetary policy announcements. Chen, Mohan and Stener (1999) examine the impact that the discount rate changes has on the equity market returns, volatility, and trading volumes using intraday data. The study found that the returns in the stock market largely react negatively and significantly to the unexpected announcements, whereas the impact of impact of expected changes on stock is insignificant. Moreover, the study also finds that stock prices react to announcements within trading hour/period after the information has been made public. This then supports the idea that unexpected changes in the discount rates impact returns regardless of the FOMC operating procedure. 2.6 Monetary Policy Transmission Channel to the Stock Market and the Real Economy It is generally accepted by economists that monetary policy changes are transmitted through the stock market via changes in the cost of capital and the wealth effect (Bernanke & Kuttner, 2005). A study by Ioannidis & Kontonikas (2008); Bjørnland & Leitemo (2009) concluded that shifts in monetary policy greatly affect stock returns and as a consequence, the idea of monetary policy transmission through the stock market is reinforced. Moreover, the quantity theory of money further qualifies the relationship between money supply as a monetary policy instrument and stock prices. When central banks increase the quantity of money in the economy, this action creates a surplus and as a result, banks lend to households and firms thus inducing demand for consumption and investment (Bissoon et al., 2016). The main link in the transmission mechanism of monetary policy can be best illustrated in the flow chart diagram below: 16 Figure 2.1 : Monetary Policy Transmission Mechanism Source: Smal & De Jager, 2001 Asset prices are affected by monetary policy outcomes through changes in the discount rate and/or changes in expected future dividends (Kontonikas & Montagnoli, 2002). It is widely assumed that equity prices are determined in a forward-looking manner, reflecting the anticipated future discounted sum of returns on the assets. Bjørnland & Leitemo (2009) argues that changes in asset prices can either be as a result of changes in expected future interest, changes in the premium of stock returns, or changes in expected future dividends. Stock valuation is achieved through future cash-flows and discounting at the appropriate interest rate level, estimated by taking into account the prevailing interest rate in the market (Bissoon et al., 2016). Earlier studies done by cook and Hahn (1989) find little evidence that market interest rates can be influenced by the decisions of the Federal Reserve (Fed). Roley and Sellon (1995) found a much closer connection but the Fed interventions and long-run interest rates appear more variable and looser. Therefore, these findings contradict conclusions made by Kuttner (2001) that changes in monetary policy have an effect on interest rates and therefore can impact discount rates. What is notable about studies that failed to distinguish that did not find monetary policy having an impact on interest rates, is that they failed to discern between anticipated and unanticipated components of monetary policy changes. The complication with approximating the response of the stock market to monetary policy sterns from the fact the stock market is unlikely to react to anticipated policy actions (Bernanke & Kuttner, 17 2001). What is interesting to observe in the literature involving the effects of monetary policy channel is the constant distinction made during announcements to capture the component of anticipated and unanticipated interest rate changes. According to Bernanke & Kuttner (2005), making a distinction to separate the effect of monetary policy into expected and unexpected on interest rates is crucial for understanding the relationship between policy actions and market interest rates. Kuttner (2001) posits that the anticipated and unanticipated effect of monetary interventions is different, he explains that the anticipated change in monetary interventions is small, whereas a big change can be observed on the unanticipated shift. The majority of economists agree that real economic activity relates to the stock market performance, whilst a few claim that stock returns play any role beyond just serving as an indicator for anticipated future corporate profits (Aziza, 2010). Bonga-Bonga (2012), highlights that equity prices affect real economic activities through various channels (consumption wealth channel and balance sheet channel). In the context of the consumption wealth channel, it relates to households owning shares in the stock market listed entities. Whenever stock prices appreciate in value, households feel wealthier as this increases their financial wealth resulting in higher consumption. However, Neri (2004) has argued that empirical evidence concluded in the US, Canada, and the UK shows that the relationship between financial wealth and consumption in reaction to changes in stock prices is not impressive. Although the effects on aggregate demand and the output passing through this channel may be small in these countries, it does not invalidate the fact that stock price increases may create incentives for consumers to take effect. The balance sheet channel effect presents itself in the form of investment decisions taking into consideration the quality of the company’s balance sheet. A rise in asset value increases the value of shareholder’s equity and thus making it favourable to raise debt funding by offering quality collateral. 18 2.6.1 Empirical evidence from the Developed Economies Schrey & Hafdísarson (2017) examined the Icelandic stock market reaction to changes in the interest rate and also to determine if the stock market was efficient between 2009 and 2017. The study was conducted using a constant mean return model and market model to estimate expected returns. Linear regression was also applied to estimate the effects that unanticipated interest rates have on stock prices. The results show that on announcement day, the anticipated interest rate changes do not affect stock prices. However, the study further demonstrated that changes in unanticipated interest rates have a statistically significant impact on stock prices on the day of the announcement. Schrey & Hafdísarson (2017) also found that the Icelandic stock market is efficient when interest rate changes are incorporated, arguing for semi-strong market efficiency. Ioannidis & Kontonikas (2008) conducted a study which looked at the impact that monetary policy has on stock returns in thirteen Organization for economic operation and development (OECD) countries between the period 1972 and 2002. The results indicate that monetary policy shifts significantly affect stock returns, thereby supporting the notion of monetary policy transmission through the stock market. The study further examined the contemporaneous impact of monetary policy outcomes on equity returns paying due regard to non-normality typically inherent as well as the significant co-movements on global stock markets. The study concluded that expansionary monetary policy boost stock prices remain largely robust in most sample countries. Although on aggregate, it has been proven that that shifts in monetary policy significantly affect stock returns in 13 OECD countries, the extent of the link between decisions by monetary policy authorities and stock markets reaction varies across economies (Ioannidis & Kontonikas, 2008). In another study, Li, Iscan & Xu (2010) conducted a comparative study on the impact of monetary policy shocks on stock prices between the United States (US) and Canada using structural VAR models. The study concluded that Canada’s instantaneous reaction of equity market prices to the central bank contractionary shock is insignificant and a brief response to the shock is recorded. Whilst in the United States, when a similar shock is induced on its domestic stock market, the impact of the monetary shock is large and a protracted response is recorded. 19 The economic architecture of these two countries explains the differences in results largely driven by differences in financial market openness. The findings of the study to this effect highlight that for a small open economy like Canada, unanticipated changes in the U.S federal funds rate significantly affect the forecast error variance of the Canadian stock market. This is consistent with the hypotheses anchored on that contractionary domestic monetary policy shocks in the context of small open economies would have a small negative impact on equity price (Li, Iscan & Xu, 2010). Li, Iscan & Xu (2010) expand this further by indicating that, in a small open economy, monetary policy is influenced by global interest rates and as a result, the impact on the discount rate is limited on the domestic stock market. At the same time, stock prices in small open economies are largely influenced by shocks originating from the rest of the world due to unexpected capital inflows and outflows. In the context of Europe, Bohl, Siklos & Sondermann (2008) measured the response of European Stock market returns to unexpected interest rate decisions. The study found that there is a negative and significant relationship between unexpected European Central Bank (ECB) decisions and European Stock Market's performance. This implies that the ECB successfully communicates its monetary policy stance to financial market participants. The interesting observation about the study is that it extracted the unexpected component of monetary policy through utilising EURIBOR future and EONIA Swap data as well financial market participants survey data covering opinions. Moreover, a study by Stoica & Diaconasu (2012) examining the impact of monetary policy on equity indexes in European Union countries also found that there is a long and short- term relationship between stock prices and interest rates. The study further discovered that during the economic crisis period, their long-run comovement between interest rates and equity markets is lower. For the US economy, Bernanke & Kuttner (2005) analysed the impact of changes in monetary policy on equity prices in the U.S and concluded that a hypothetical unanticipated 25 basis point cut in the federal funds rate target is related to about a 1% increase in broad stock indexes. The VAR model study by Bjørnland & Leitemo (2009) reached a similar conclusion that there is a great interdependence between interest rate setting and stock prices, whereby a Monetary shock 20 that raises the federal funds rate by ten basis points immediately pushes stock prices to fall by 1.5 percent. 2.6.2 Empirical evidence from the Developing Economies Iddrisu, Harvey & Amidu (2016) comprehensively examined the monetary policy and stock market dynamics from the African perspective over the period 1979-2013 using the VAR technique. The study found that the stock market of the 12 African countries is positively affected contemporaneously by their respective monetary policies through the interest rate channel, but was not able to find evidence of the reverse reactions. Furthermore, the study looked at the impulse response and established that both interest rates and money supply fall in response to positive and negative shocks in the stock market. The study through the forecast error variance decompositions (FEVD) tried to establish which monetary policy tool (Money supply and real interest rate) had the greatest influence on the stock market and found that real interest rate had the greatest impact on the stock prices. Ajie & Nenbee (2010) examined the relationship between monetary policy and stock prices in the Nigerian stock exchange. The study employed a co-integration and error correction modeling to conclude that both money supply and interest rate were significantly correlated with stock prices. The data sample was quite extensive which looked at the periods 1986-2008 taking into consideration the financial sector liberalization. Ali, Adeeb & Saeed (2014) studied the dynamic impact of monetary policy on stock returns focusing on the manufacturing sector of Pakistan between the period 2001 and 2010. The study reveals that monetary policy and company-specific factors have a significant impact on stock returns, thereby confirming the notion of the monetary policy transmission mechanism. In the South African Context, Gupta & Reid (2013) explored the sensitivity of industry stock returns to monetary policy and macroeconomic news through the Bayesian Vector Autoregressive (BVAR). The study concluded that monetary policy surprise is the only variable that significantly 21 and consistently negatively affects the stock market, whilst the Consumer Price Index (CPI) surprise plays a significant role. Another study by Muroyiwa, Ezeoha, & Mushunje (2017) estimated the relationship between SARB’s monetary policy decisions and JSE performance instrumenting with Structural Vector Autoregression (SVAR). The study concluded that a monetary policy announcement that increases the interbank rate by 100 basis points result in a decrease of 1% in the S.A stock returns. This however shows that the strength of the link is moderately minimal for S. A compared to other countries in the developed world. Furthermore, a study by Mangani (2011) looked at the effects of monetary policy on JSE portfolios using a Generalised Autoregressive Conditional Heteroskedasticity (GARCH) model which covered periods 1990-2009. The results revealed that changes in the discount rate are very crucial in describing the mean returns and return volatilities of the JSE portfolios. Through the Structural Vector Error Correction (SVEC) model, Bonga Bonga (2009) investigated the interconnection between monetary policy, equity prices, and economic activities on emerging markets using South Africa as a case study. The study found that the impulse response functions display that, the response of expected inflation to equity price shocks is negative for the first two quarters before it becomes positive and statistically significant in the seventh quarter. The findings are consistent with a study done by Chinzara (2010) to investigate the effect of macroeconomic volatility on equity markets volatility using the multivariate Vector Autoregression Model. The study found that there is a negative volatility spillover from inflation and this can be attributed to the possible structural breakdown of the relationship between inflation uncertainty and stock market volatility. This was due to the introduction of the inflation targeting policy framework by the South African Reserve bank (SARB). Mallick & Sousa (2011) extended the study on the monetary policy transmission to cover the fastest-growing emerging market economies: Brazil, Russia, India, China, and South Africa (BRICS). The monetary policy (interest rate) shocks are identified using modern Bayesian Methods and the panel Vector Autoregression framework and the study found that contractionary monetary policy has a strong and negative impact on equity markets. The panel VAR exercise 22 provided further robustness on the findings that contractionary monetary policy harms output for this group of key emerging market economies. However, Adam & Tweneboah (2008) established that there is a positive relationship between monetary variables such as inflation and stock price returns in Ghana. The study looked at the role of macroeconomic variables in the movement of stock prices between 1991 and 2006 by exploring the long-run relationship between the variables utilizing Johansen’s multivariate cointegration tests. The short-run dynamics were explored through the impulse response function and forecast error variance decomposition analysis. Moolman & Du Toit (2005) developed a structural, theoretically-founded model for the South African stock market estimated using co-integration and error-correction techniques between the period 1993 and 2003. The study found that short factors such as interest rates, the rand-dollar exchange rate, and the S&P 500 index influenced the short-run movements, while the discounted dividends determined the long-run movements of the stock market. 2. 7 Application of Forward Rate Agreements The standard view of the monetary transmission mechanism relies on a simple version of the expectations theory of interest rates (Roley & Sellon, 1995). This theory implies that long term rates are an average of short term rates and expected future short term rates. Roley & Sellon (1995) further illustrated on this theory by explaining that an increase in the desired level of federal funds rate causes current short term rates and expected future short term rates to rise, which ultimately pushes interest rates across all maturities. Central banks and market agents of ten find the forward interest rate market to be of greater interest since it states explicitly expectations of future interest rates and thus, Forward rate agreements can serve as an indicator for the term structure of interest rate (Malz, 1998). He further argues that FRAs can communicate a message about the consensus of future short term rates and the near term policy position of the monetary policy. Mwaza (2020, p.6) defines Forward rate agreements as an “over the counter agreement to earn or pay an interest rate on a deposit starting at a future point in time”. This simply means that two parties agree to set future borrowing rates in advance. It is often argued in the simple efficiency specifications of forward exchange market that, forward rate will fully incorporate all available 23 information about the future rate expectations (Chiang, 1988). Therefore, this means that forward rates can be used to predict the spot price of interest rates in an unbiased manner as long as market participants are able to process information rapidly. Wesso (1999) posits that through the arbitrage activities of economic players and market adjustments, the forward rate market reflects information that is expected to predict future spot rates or the future discount rate that can be used in stock valuations. In determining monetary policy, central banks must take into account relationships that exist spot exchange rates, forward rates and interest rate margins (Wesso , 1999). A disruption of these relationships could easily lead to new speculative transactions in the stock exchange market. FRA prices are expressed as the rate the buyer pays on the notional deposit, that is, as a forward interest rate. FRAs are cash settled, by the present value of the difference the realised short term rate and the FRA price, so no deposit will actually be made (Malz, 1998). The quote on FRAs is the risk neutral market estimate of the future money market rate. A standard FRA involves three points in time, (a) the current time (b) the expiry time 𝑇𝑖−1, ( c) the maturity time 𝑇𝑖, with t≤ 𝑇𝑖−1 ≤ 𝑇𝑖 . At any point in time s ∈ [𝑡, 𝑇𝑖−1], the fair value of a FRA with a FRA rate of K to the buyer is given by: 𝑉𝐹𝑅𝐴 = 𝑁𝜏 (𝑇𝑖−1, 𝑇𝑖) [𝐿 (𝑠; 𝑇𝑖−1,𝑇𝑖) − 𝐾] 𝑍 (𝑠, 𝑇𝑖 ) (1) Where:  N is the notional amount of the trade;  𝜏 (𝑇𝑖−1, 𝑇𝑖) is the difference between time points 𝑇𝑖−1 and 𝑇𝑖 in year fractions ;  L (𝑠; 𝑇𝑖−1,𝑇𝑖) is the simple fair forward rate at a time for the period[𝑇𝑖−1, 𝑇𝑖];  Z (s, 𝑇𝑖) is the discount factor from time s to time 𝑇𝑖 In the market, FRAs are denoted by short hand notation like 3x6. This refers to a FRA with start date 3 months from now and maturing 3 (6-3) months later. Thus, a 3x6 FRA is contract fixing the 3-months JIBAR rate in 3 months’ time. 24 Although FRA’s are over the counter instruments in South African as it does not have a futures market to trade repo rates like it is the case in the U.S that has a well-functioning futures market, the principle is well aligned. According to Malz (1998) FRAs indicate expected future interest rate at a specific horizon which aid in monitoring market expectations of the interest rate on or around a specific date and this make them more preferable also to monitor expectations of the pace of, for example, interest rate changes over the next 12 months. 2.8 Overview of the South African Monetary Policy in last 10 years With the ever-changing globalised economy, the South African monetary policy has become increasingly exposed to a variety of challenges in its efforts to achieve domestic price and financial market stability (Smal & De Jager, 2001). When the reserve bank decides to intervene in the market by employing monetary tools at its disposal, it stimulates a series of economic events. Monetary policy in South Africa is executed by the South African Reserve Bank (SARB) which derives its mandate from the constitution of the republic. Section 224 of the constitution of South Africa mandates the SARB to achieve and maintain price stability in the interest of balanced and sustainable economic growth. Although inflation targeting by the SARB is its primary concern, it also attempts to steer the economy off recessions, stabilises the rand or output fluctuations, as well as other financial instabilities. The SARB must perform its functions independently and without fear, favor, and prejudice. Whilst independent in its operations, the SARB is further mandated to have regular consultations with the Minister of Finance to ensure that both fiscal and monetary policy are well aligned to achieve the stability of the financial system. The responsibility of the SARB also extends to accumulating foreign currency reserves and consequently, it will venture into the foreign currency market although it not directed by the constitution to follow the movements of currencies in the exchange rate market (Ramatlo, 2019). Repo rate changes have a short impact on the exchange rate. The monetary policy of S.A in the past 10 years have been overseen by two governors: the ninth governor, Ms. Gill Marcus, and the 10th Governor, Mr. Lesetja Kganyago. Both serving under the government of the African National Congress (ANC). The monetary policy regime pursued under the two Governors has been anchored around inflation targeting adopted in February 2000. This 25 move was critical to repositioning the SARB amongst world central banks as a credible and reputable institution as inflation targeting enhanced policy transparency, accountability, and predictability. A Stable inflation environment is one form of growth-enhancing certainty that is enabled by greater predictability of both monetary and fiscal policy. For example, transparent and effective inflation targeting will guide inflation path and constraints (Aron & Muellbauer, 2005). A clear inflation path creates a conducive environment for the private sector to plan on future expenditure and investment which also makes it easy for wage settlements and pricing of goods and services to take effect. The sampled period was further characterised by a post-recession recovery period which recorded a wide range of monetary responses across the globe. There was a convergence from developed economies central banks to engage in a wide-scale quantitative easing to stabilise financial markets and to arrest the decline in inflation, growth, and employment induced by the global recession. The recent global economic downturns have placed monetary policy interventions in the spotlight. It is now argued by economists that monetary policy should be deployed as the first line of defense to stabilise the economy during economic slowdowns (Matemilola, Bany-Ariffin, & Muhtar, 2015). However, the rate at which economic stability can be realised depends on the pass-through to the bank lending and how developed the capital markets are. The South African monetary policy in the last 10 years has made significant strides in stabilising prices by maintaining inflation within the target range of 3-6% as observed in figure 2, underpinned by the stability of the financial system and financial markets. The SARB uses different policy tools such as the, reserve requirement, liquidity requirement, repo (repurchase rate) rate etc. in its attempt to effectively maintain price stability. The repurchase rate is defined as the rate at which the commercial banks borrows money from the SARB and therefore represents a cost of credit to the banking sector (Matemilola, Bany-Ariffin, & Muhtar, 2015). The mandate of the SARB of stabilising prices within the target range of 3-6% is executed by changing the repo rate. To illustrate, when inflation moves close to the upper or even breaches the 6% band, the reserve bank adjusts the repo rate higher to pull inflation back to the target range. The repo rate is adjusted lower when inflation moves in the opposite direction to the lower band. The monetary policy outcomes are communicated by the Governor of the bank on behalf of the Monetary Policy Committee 26 (MPC) meeting. The SARB take the local and international audience into confidence by explaining the reasons behind their decision(s) about adjusting repo rates differently in response to inflation. Table 2.1 MPC scheduled and Unscheduled Meetings Period Sample Period: 2010-2019 Total no. of scheduled meetings per year 6 Total no. of unscheduled meetings per year 0 Total no. of MPC meetings for sample period 60 Increase in repo rate 7 Decrease in Repo rate 7 Unchanged 46 Data Source: SARB MPC Policy Statements, 2010-2019.Table constructed by author Table 2.1 demonstrate that the South African Monetary Authorities settled on 6 fixed meetings per year which have been the case since 2010. Period 2004 to 2008 scheduled MPC meetings varied between 2 to 3 whilst 2009 had about 9 meetings. This was necessary due to the rising risks posed by the Global recession on the South African economy. There were no unscheduled meetings in 10 years under investigation. Unscheduled meetings are normally called to respond to an emergency event that has a material impact on inflation or the South African rand. Therefore, this paints a positive picture of the stability of inflation within the required target range that did not necessitate the SARB to call extra-ordinary meetings during the period. The above table further shows that 77% of MPC meetings voted to leave repo rates unchanged, meaning that the risk of inflation was perceived delicately balanced. The remainder of the meetings (23% of MPC meetings) adjusted repo rates equally to respond to the downward and upward risk on inflation. 7 meetings voted to increase interest rates, whilst the other 7 voted to decrease interest rates. From table A3 in the appendix, it can be concluded that when the SARB does change the repo rate, the changes are usually between 25 basis points and 100 basis points. 27 Figure 2.2 SARB Repo Rate (%) Data source: Bloomberg. Graph plotted by author Figure 2.2 shows us the repo rate during our sample period. Successive rate cuts brought the repo rate from its high of 7 percent in the year 2010 to the lowest point of 5 percent around mid-2012. A gradual tightening cycle began lifting the rates to around a 7 percent level beginning of 2016. This is the high recorded at the beginning of the sample period (2010). In general, monetary policy easing occurs when economic growth is slowing, while monetary policy tightening occurs when economic growth starts picking up. 28 Figure 2.3 Inflation and GDP Growth in (%) Data source: Bloomberg. Graph plotted by author From Figure 2.3, we can see that the economy show signs of slowing from the end of 2010, with real growth declining steadily by the end of 2013. A short rebound from negative growth was observed in the first quarter of 2014 but the economy has experienced a sideways growth until the end of the sample period. What is interesting to observe from figure 2.3, inflation faired above SARB’s midterm point of 4.5 percent for most of the sample period disregarding the sideway movements of economic growth. Although the South African monetary policy targets inflation and made price stability its sole objective, data on a range of macroeconomic indicators such as real economic output and unemployment shows that the attempt by the SARB to influence these variables is muted. Figures 2.2 and 2.3 shows that while the SARB was continuously cutting rates between 2010 and 2013, economic growth was declining until the South African economy registered a negative growth of 0.6% in the first quarter of 2014. There are several explanations for the persistent decline and unexpectedly sluggish recovery of the South African economy during the sampled period which cannot be blamed squarely on the inadequacies of monetary policy to boost economic growth. For 29 example, there is a high level of policy uncertainty that undermines confidence in the South African economy, and the lack of political will to implement reforms contained in the National Development Plan (NDP), high regulatory costs, corruption both in the public and private sector etc. These factors are some of the big contributors that undermine economic activity and may impair the monetary policy transmission channel from effectively stimulating output. Figure 2.4 Unemployment Rate in (%) Data source: STATS SA (2010 to 2019). Graph plotted by author Employment is a key macroeconomic variable in which monetary policy objectives are measured. From figure 2.4, it can be observed that unemployment held steady at around 25.5% from the beginning of the sampled period until it started rising rapidly in mid-2013. The rising period of unemployment coincides with the rapid decline in economic growth as observed in figure 2.3. What is notable about the unemployment trend before it started rising end of 2013, the period of stability was supported by successive cuts in repo rates until the end of 2013 as observed in figure 2.2. As repo rates started to increase in response to the steady rise in inflation, unemployment started to creep up. 30 2.9 Chapter Summary The literature in this study provides a broad background of how monetary policy can directly impact financial assets and how this relationship can feed into the real economy. The channel in which monetary policy affects the stock market has been fully explored by examining empirical studies both in the developed and developing world. The majority of studies were able to establish a relationship between monetary policy outcomes and the stock market, however, very few were unable to trace this link. It was uncovered that those studies which were unable to observe a link between monetary and asset prices did separate the anticipated and unanticipated component of monetary interventions. Furthermore, a background of how central bank communications can shape and manage expectations in the financial markets was explored. Due to the linkages between Central bank communication and transparency, it was important to look at this component in detail as it can stabilise financial markets. Literature also shows that many researchers have employed the event study by instrumenting with a VAR methodology to examine this relationship (see for example Bernanke and Kuttner, 2005; Bjørnland & Leitemo, 2009; Iddrisu, Harvey & Amidu (2016)). And all the studies mentioned above found a negative relationship between an interest rate shock and stock returns. Bredin, Gavin & O'Reilly (2003) used an OLS (Ordinary Least Squares) method to arrive at similar conclusions to the researchers above. This shows us that there several methodologies to conduct this kind of a study. Also, many considerations and assumptions were made to analyse the relationship between monetary policy interventions and the equity market. This presents an interesting area of study to explore further methodological approaches to arrive at robust conclusions that is consistent with theory. Lastly, literature also looked at the development of monetary policy in South Africa in the past 10 years. Literature delved deeper to identify the mandate under which the SARB operates under which is used to attain a set of objectives. South African monetary policy mandate is clearly defined in the constitution and has given the SARB powers to stabilises prices. Within the sampled period under investigation, the SARB was successful in maintaining price stability. 31 Chapter 3. Data Description and Research Methodology 3.1 Introduction Babbie and Mouton (2001) define research methodology as methods, procedures, and techniques used in the process of the research plan and design implementation. This chapter explores the conceptual framework on the impact of monetary policy outcomes on the South African stock market. It also took a critical look at the methodology for assessing the impact of monetary policy variable(s) on the stock market, specified the model, and diagnostics tests were done to determine the precision of the model. 3.2 Research Design Research design is a plan that maps out the process which the writer followed to clarify the research objective outlined in section 1.4. Williams (2007) notes that the research process is a systematic attempt to define the objective, organising and managing data, and presenting the findings. The first step followed to gather preliminary information on the impact of monetary policy announcements on stock returns was through reading published empirical journals, books, and other relevant papers. This study also looked at literature on various economic jurisdictions as well as the importance of stock market development to ascertain the variables to be used. This study utilised secondary data between the period January 2010 and November 2019 sourced from the Bloomberg terminal which provides credible digital data and news for financial, economic institutions, and other market participants. This study further used descriptive statistics to sufficiently explore the characteristics of the variables. 3.3 Population and Sample The ultimate objective of quantitative research is to generalize findings, as it is impossible to study the entire population of the research interest (Khalid, Abdullah & Kumar, 2012). Therefore, this study identified a defined population and sample that was instrumental in generalising the research findings. Sample represents a subgroup of a population, therefore allowed this study to draw 32 inferences about the population. The population is defined as a complete set of units (persons or objects) that displays some common characteristics defined by the sampling criteria established by the researcher (Molenberghs, 2010). This study followed a non-probability and purposive sampling technique. Purposive sampling also called a judgmental or subjective sampling is described as deliberate choice made by a researcher to identify a representative sample with qualities and characteristics similar to the population of the study (Sharma, 2017). The sample in this study is made up of the Johannesburg Stock Exchange (JSE) all-share index, JSE Financial Index, and the JSE Industrial Index. The JSE All-share index was selected because it represents a total (broad-base) index, reflecting a total picture of the behaviors of the common quoted stocks on the South African Stock Exchange. The Sub-indices used in this study were selected to further examine the impact of monetary policy on stock returns at the sectoral level. Figure 3.1 Plot of Stock Market Variables Data source: Bloomberg. Graph plotted by author Figure 3.1, shows that the stock market represented by the JSE ALL Share, Industrial Sector, and the Financial Sector had modest growth and never returned to the early recovery periods of 2010. It encompasses the allure of high yields that attracted investors with a higher risk appetite to emerging markets like South Africa, contributing to the faster recovery and the stability of the JSE. 33 This study employed time series percentage change daily closing prices of the South African stock market (Johannesburg Stock Exchange All Share Index, sub-indices), actual alterations in the SARB repo rate, and the FRA (Future Rate Agreement) derived from Bloomberg terminal, spanning the period January 2010 through to December 2019, which looked at the post-global recession dynamics. This period was considered to shed light on the impact of monetary policy announcements on stock prices because it covers a period where many economies especially those with heavy natural resource dependence, such as South Africa, experienced a slowdown in economic growth and a decline in many economic activities. FRAs were considered and used in this study because they indicate expectations of changes in future policy. Ioannidis & Konotonikas (2006); Bernanke & Kuttner (2005) used federal futures data as the official monetary policy rate representation. Whilst in the South African context, Ramatlo (2019) used a one month Johannesburg Interbank Agreed Rate (JIBAR). Figure 3.2 below shows the movement of the 2 x 5 movement of the FRA in relation to the repo rate during 2010- 2019 period. Kuttner (2001) argues that the advantage of using forward rate data as a measure of expected monetary policy; there are no problems with model selection, no generated regressor issues and lastly, the out-of-date to produce the estimates are a non-issue. This embodies short- term expectations of the SARB repo rate, as it offered a promising way to measure the unanticipated element of specific MPC actions. The other advantage FRA 2x5 brought into this study, its price is based on a specific day of the month and it gives a correct measure of the expected repo rate. This study considered a short duration FRA to properly evaluate expectations of monetary policy. MPC policy announcements are six weekly, the 2x5 FRA is preferred over other shorter rates (1x4 and 3x6) as it effectively incorporates expectations for the future monetary policy rate as meetings rarely exceeds 2 months. The 1x4 FRA is a 3-month interest rate in one- 34 month time and therefore does not sufficiently cover the period, while the 3x6 incorporates expectations of two meetings. Figure 3.2 Plot of Repo Rate and Fras 25 Data source: Bloomberg. Graph plotted by author 3.4 Data Collection Data used in this study was collected from Bloomberg Terminal. The data collected is daily closing stock price data for the JSE all-share index, Repo rates, and Forward Rate Agreements (FRA’s). The daily data covers period: Jan 2010 to Nov 2019. The period was identified to enable the study to assess the post-recession dynamics, also most studies in the South African context by (Gupta & Reid, 2008; Mangani, 2011; Bonga Bonga, 2009) looked at data before the Global Financial Crisis of 2007/08. 3.5 Measurement of Variables A researcher must have a proper understanding of the nature of variables under study and how to measure it (Khalid, Abdullah & Kumar, 2012). In this study, the dependent variable used is the (JSE all share index), whilst the independent variables were Forward Rate Agreement (FRA) and 35 Repurchase Rate (REPO). Although the two variables are easily identifiable, the researcher also paid attention to extraneous variables or often called experimental errors which may not be related to the study but could affect the dependent variable. 3.6 Data Analysis The section below describes the data analysis approach in detail 3.6.1 Event Studies Bredin, Gavin & O'Reilly (2003) used the event study methodology to measure the response of the equity returns to unexpected changes to monetary policy outcomes on the day of the announcement. However, the market may also react to the lack of change in the repo rate if there was a market-wide expectation for a change (Bernanke & Kuttner, 2005). According to Bredin, Gavin & O'Reilly (2003), the appropriate identification of monetary policy changes and the need to discriminate between expected and unexpected changes in policy interventions have been the most methodological considerations that have influenced this type of study. The event study assumes that scheduled monetary policy announcements are the dominant influence on the level of stock prices on the day of the announcement and the day immediately after the announcement. The event study method was carefully selected to try to control the influence of other variables or information outside of monetary policy announcements. This was achieved by using high-frequency data (Daily) which enabled the study to examine a narrow time interval surrounding the policy action. In particular, this study chose the event data which represents the announcement date (t) and a day preceding (T-1) and succeeding (T+1) the announcement day, as captured in figure 3.3, event window. Therefore, the changes in the level of stock prices can largely be attributed to this announcement. An event study examines the return behavior for a sample of stocks experiencing a common type of event. Given the above explanation, the event might take place at different points in calendar time or it might be clustered at a particular date (e.g., African National Congress elective conference 36 outcome affecting an industry or a subset of the population