First, an attempt has been made to estimate the volatility of spot and derivatives, i.e. futures markets separately in different modeling framework. Initially, time invariant measure of volatility tests like standard deviation, skewness and kurtosis are derived and compared for both the markets amongst different asset classes. Since it is a well established fact now that the time varying nature of volatility can be well captured in ARCH family of models, an effort has been made to estimate the volatility of both spot and futures markets utilizing different ARCH family of models. Then the forecasting power of different models are tested through different forecasting measures in order to find out the model with lowest forecasting error or maximum predictive accuracy. By comparing the return volatility in spot and derivatives markets, it is possible to find out whether the spot market possesses higher volatility compared to the derivatives market or vice versa.
Thus, the present research is being conducted with the following specific objective :
To test the volatility of spot and futures markets separately and compare. Apart from measuring volatility, we also tried to find out the best volatility forecasting model among different ARCH family of models utilized.
Therefore the hypothesis, attempted to be tested for this specific objective are:
i) Volatility in spot and futures markets in India are indifferent, and
ii) Different ARCH family of conditional volatility models are equally accurate and significant in forecasting the volatility of underlying indices and stocks, both in spot and futures market.
DATA FOR MEASURING VOLATILITY OF RETURNS
Data used to test the volatility of return in both futures and spot markets are daily closing prices of assets traded and quoted at National Stock Exchange, Mumbai. The indices used here include NSE S&P CNX Nifty cash and futures index. The stocks selected for the purpose are RELIANCE INDUSTRIES, INFOSYS, HINDUSTAN UNILEVER, HDFC, HINDALCO, ACC, TISCO, L&T, SBI and TELCO. These are high turn over, high profit making blue chip stocks included for computation of popular sensitive indices like BSE Sensex and NSE CNX S&P Nifty representing diverse sectors of broad economy and continuously traded both in cash and futures markets of stock exchanges and provide enough liquidity to the system. As far as the frequency of index and stock data is concerned, it includes only daily data. Logarithmic returns are calculated from the daily closing price observations over a sample period starting from 1st January 2002, i.e the year after initiation of futures trading till 31st December 2010. The stock futures returns for near month or one month contracts are only taken into consideration. Data on the indices and underlying stocks have been collected from the NSE web site (www.nseindia.com). All the time series data are adjusted for non-synchronous trading effect, if any. The volatility measures for index and ten underlying blue chip sensex stocks are conducted within this sample period.