Interaction amongst Equity, Forex and Bond markets: evidence from Singapore.

• Author: Chakraborty, Sandip

1. INTRODUCTION

   Interdependence among financial markets has received significant attention in the finance literature. Understanding the behavior and sources of financial markets' linkages is important for diversifying, pricing securities and asset allocation decision making. The volatility in the finance markets depends upon numerous macro-economic factors that have different degrees of importance to different economies of the world. There have been studies conducted for the purpose of analyzing inter-relationship amongst various macro-economic variables. Pan et al (2007) analyzed the dynamics linkages between exchange rates and Stock market for seven east Asian countries and found that a causal relation exist from the equity to forex market. Kim (2003) studied the relationships among stock price, real exchange rate, interest rate, and inflation in the United States and found Stock price has negative relation with all other factors. In contrast, Balduzzi (1995) conducted a study on stock return and inflation and found weak negative correlation between them. Najand and Noronha (1998) investigated the causal relations among stock returns, inflation, real activity, and interest rates for Japan and found that inflation had negative impact on Stock returns and explained dynamics of interest rates. Gjerde and Saattem (1999) conducted a similar study in Norwegian market; small open economy, and found that real interest rate changes affect both stock returns and inflation. Chan, Norrbin and Lai (1997) studied the collinearity of stock and bond prices are and found that they are integrated of the first order. On the contrary, Kim and Francis (2007) investigated the relationship between stock prices and bond yields in G7 countries and found that stock prices and bond yields do not move together. Borensztein et al (200 (1)) analyzed the impact of short run effects of US interest rate shocks and EMBI1 on various economies and found US interest rate shocks led to a greater impact of EMBI as compared to the shocks led by domestic interest rates of respective economies.

   Panchenko and Wu (2009) recently investigated the extent of stock market integration affecting the joint behavior of stock and bond returns using sample from eighteen emerging markets and found robust inverse relationship suggesting that stock market openings lead to an increase in demand for equities and an unchanged or reduced demand for bonds. Fleming et al (1998) investigated the nature of volatility linkages in the stock, bond, and money markets and found the relationship has evolved from strong to stronger after 1987 market crash. There is a stream of available literature aiming to analyze correlation of Volatilities of stocks, currencies and bonds (for example Tai, 2000; Elyasiani et al, 2004; Tai, 2007; Tse et al, 2002; Alaganar et al, 2003).

   Monetary policies in Singapore have been centered on the management of exchange rate. Singapore dollar is managed against an undisclosed trade weighted basket of currencies basket of foreign currencies and the policy is reviewed in a three months cycle. Studies conducted earlier (Chow, 2007) provide a detailed description of Singapore's monetary transmission mechanism which suggests that the exchange rate is more influential than the interest rate as a source of macroeconomic fluctuations. Studies conducted by IMF (Parraddo, 2004) have analyzed the reaction of inflation and TWI (2) by deriving a model using six variables to describe the Singapore economy: industrial production, the CPI inflation, a world commodity price index, a foreign interest rate, a domestic interest rate, and the TWI. Publications from Monetary Authority of Singapore (Khor et al, 2007) describe a reaction function of the Singapore currency. Gali & Monacelli (2002) analyzed the exchange rate volatility in small open economy using domestic inflation and CPI and an exchange rate peg. There are studies conducted on the relationship of stocks and bonds reaction of Asian markets and Singapore (see. Tang and Shum, 2004; Maysami and Koh, 2000; Panchenko and Wu 2009). However there are not many studies which analyzed the relationship of Stocks, bond, forex markets and interest rates in Singapore; especially in the context of structural breaks and appropriate model specification. Hence there lies an opportunity to analyze whether there exist a relationship between these markets in Singapore and how strong or weak the relationship is. Singapore Overnight Rate Average (SORA) was introduced by MAS in July 2005. SORA is an index that tracks actual Singapore Dollar overnight funding rates transacted by market participants. Albeit the economic significance of SORA, current research seems to have neglected the dynamics of SORA vis-a-vis other markets (Stocks, Bond and Forex) in the context of structural breaks.

   Considering the economic position of Singapore and the market volatility due to recession, it would be interesting to analyze how Stocks, Bond, Forex markets and interest rates are related. From a portfolio risk management perspective it is important to know how market moves and which markets moves together and which doesn't and forms a good area to be analyzed. The correlation between these markets could be time varying and will impact the portfolio risk & returns. Fundamental pillars of financial econometrics are predicting the level of uncertainty and the strength of co-movements of asset returns (Engle, Focardi and Fabozzi, 2007). Dynamics of interest rates and exchange rates and transmissions are significant in the management of Current assets and liabilities.

2. OBJECTIVE OF STUDY

   Objective of this study is to identify possible structural breaks within Singapore Overnight Rate Average (SORA) and further analyze the volatility of SORA within break periods if any. This study also attempts to study the volatility and correlation of SORA with Singapore Strait Times Index, Singapore Dollar-US Dollar rate, Short Term Debt Securities (1 year), T-Bills, Singapore REPO Rates, Federal Fund Rates, S&P 500 Index, Morgan Stanley Capital International (MSCI) Index, and Debt Security Yields w.r.t structural breaks in SORA.

3. DATA AND METHODOLOGY

   The required data was collected for the purpose of this study from various sources as listed bellow: Data Series Source Singapore Economic Indicators Statistics Singapore Website (3) Business Cycle information NBER (4). Singapore Stock Indexes ( STI) Yahoo Finance (5) SORA MAS website (6) SGS Overnight REPO MAS Website SGS Prices and Yields MAS Website (3M, 1Y,2Y,5Y,7Y,10Y) USDSGD Forex rates Federal Reserve website (7) S&P500 Yahoo Finance MSCI--The World Index MSCI Barra (8). website (9) standard core BBA REPO Rates BBA Libhor website (10) US Federal Funds Rate Federal Reserve website US Treasury Constant Maturity Rate Federal Reserve website (3M, 1Y,2Y,5Y,7Y,10Y) The Business Cycle information has been collected from NBER database. According to NBER, the period from July 2005 till Aug 2009 can be categorized as Non-recessionary (03 Dec 2001 to 31 December 2007) period and as Recessionary (01 January 2008 to 31 August 2009)

   Once the data was collected, cleansed and segregated into Recessionary and Non-recessionary subperiods, the returns for each data series, viz. SORA with Singapore Strait Times Index, Singapore DollarUS Dollar rate, Short Term Debt Securities (1 year), T-Bills, Singapore REPO Rates, Federal Fund Rates, S&P 500 Index, Morgan Stanley Capital International (MSCI) Index, was computed as logarithm of ratio of current price to

   previous day's price. Since objective of this study is to find out the degree of interaction between the above said indices, the choice of using Multivariate GARCH model was an obvious one. However, such a model can be biased if one ignores structural break analysis. The parameters of Multivariate GARCH do not assure stability over a period of time unless possibility of structural breaks is ruled out. Therefore, Structural Break test was carried out w.r.t SORA to test for the parametric stability of time series data of SORA. Upon examining about the possibility of Structural Breaks with SORA, Multivariate GARCH models were tried to be fit with a view to estimating the degree of...