Microsoft Word - 4830-17681-2-SM-writer2-new.doc Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 423 A Study on Options Pricing Using GARCH and Black-Scholes-Merton Model Zohra Bi Department of Finance, School of Business, Alliance University, Bangalore, India E-mail: zohrayousuf@gmail.com Abdullah Yousuf Department of HR, Jain University, Bangalore, India E-mail: yousufaroma@gmail.com Mihir Dash Department of Quantitative Methods, School of Business Alliance University, Bangalore, India E-mail: Mihir@alliance.edu.in Received: Dec. 24, 2013 Accepted: June 17, 2014 Published: June 17, 2014 doi:10.5296/ajfa.v6i1.4830 URL: http://dx.doi.org/10.5296/ajfa.v6i1.4830 Abstract Options are instruments which have the special property of limiting the downside risk, while not limiting the upside potential, thus their use in hedging. The share of the options market in the Indian capital market has increased to 64% in just over a decade. The trading turnover of options in the FY11 was Rs. 193,95,710 crore, and the trading volume generated by options market was almost two times that of the volume generated in the cash market and futures market put together. So trading and pricing of stock option have occupied an important place in the Indian derivatives market. Volatility is a critical factor influencing the option pricing; however, it is an extremely difficult factor to forecast. Hence the crucial problem lies with the accurate estimation of volatility. The estimated volatility can be used to determine future prices of the stock or the stock option. Empirical research has shown that using historical volatility in different option pricing models leads to pricing biases. The GARCH (1, 1) model can be a solution for this Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 424 problem. The present study applies the GARCH (1, 1) model to estimate the volatility, and applies this estimated volatility to calculate option prices with the help of Black-Scholes-Merton model. Keywords: GARCH model, Black-Scholes-Merton model, Option prices, Volatility Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 425 1. Introduction An option is a derivative financial instrument that specifies a contract between two parties for a future transaction on an asset at a reference price called the strike price. The buyer of the option gains the right, but not the obligation, to engage in that transaction, while the seller incurs the corresponding obligation to fulfill the transaction. In return for assuming the obligation, called writing the option, the originator of the option collects a payment, the premium, from the buyer. So the loss for an option buyer is limited to the premium paid, whereas the loss for an option seller is unlimited. Many options are created in standardized form and traded on an options exchange among the general public, while other over-the-counter options are customized ad hoc to the desires of the buyer, usually by an investment bank. The price of an option derives from the difference between the reference price and the value of the underlying asset plus a premium based on the time remaining until the expiration of the option. There are two types of options: call options and put options. A call option conveys the right to buy the underlying asset at a specific price, while a put option conveys the right to sell the underlying asset at a specific price. Option contracts have the following specifications: the type (call or put), the quantity and class of the underlying asset, the strike/exercise price (i.e. the price at which the underlying transaction will occur upon exercise of the option), the expiration date (the last date the option can be exercised), and the settlement terms (for instance, whether the writer must deliver the actual asset on exercise, or may simply tender the equivalent cash amount). Also, there are two option styles: European style options can be exercised only on the expiry date, while American style options can be exercised any time before the expiry date. The Black-Scholes-Merton model (1973) is the most widely-used model of determining option prices. The model expresses the prices of European call and put options on a non-dividend-paying stock in terms of five parameters: the spot price of the underlying stock, the exercise price at which the transaction will be executed, the expiration period after which the option can be exercised, the risk-free rate of return, and the volatility of returns of the underlying stock. Volatility is a critical factor influencing the option pricing; however, it is an extremely difficult factor to forecast. Hence the crucial problem lies with the accurate estimation of volatility. The estimated volatility can be used to determine future prices of the stock or the stock option, and thus an investor can use arbitrage strategies accordingly to benefit from the model. 2. Literature Review There is a vast literature on options pricing using the GARCH-Black-Scholes-Merton model. Some of the relevant literature is reviewed in the following. Adesi et al (2007) proposed a method for pricing options based on GARCH models with filtered historical innovations. They found that their model outperformed other GARCH pricing models and Black-Scholes models empirically for S&P 500 index options. Their model was validated by empirically obtaining decreasing state price densities per unit probability. Also, their model explained implied volatility smiles by the negative asymmetry of the filtered Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 426 historical innovations. The study also provides empirical evidence and quantifies the deterioration of the delta hedging in the presence of large volatility shocks. Cristofferson et al (2004) extended their results in the presence of conditional skewness. Siu et al (2004) proposed a method for pricing derivatives under the GARCH assumption for underlying assets in the context of a dynamic version of Gerber-Shiu's option-pricing model. Instead of adopting the notion of local risk-neutral valuation relationship (LRNVR) they employ the concept of conditional Esscher Transforms to identify a martingale measure under the incomplete market setting. Under the conditional normality assumption for the stock innovation, the pricing result is consistent with that of Duan. In line with the Gerber-Shiu's option pricing model, they also justify the pricing result within the dynamic framework of utility maximization problems which makes the economic intuition of the pricing result more appealing. Numerical results for the comparison of the model with the Black-Scholes-Merton option pricing model are also presented. Dash et al (2012) applied the GARCH options pricing model for options traded on the National Stock Exchange, India. They used the GARCH(1, 1) model to obtain volatility projections, and calculated option prices using these volatility projections in the Black-Scholes-Merton model. They found that the implied volatilities (for both calls and puts) were overestimated, and that call and put option prices were predominantly overvalued, and, further, that put options were more overpriced than call options. They also found that the overestimation of volatility and overvaluation of options prices increased with higher market capitalization and moderate/higher trading volume of the underlying stocks. Duan (1995) introduced the GARCH option pricing model, linking econometric models with the options pricing literature. Heston and Nandi (2000) developed a closed-form option valuation formula for a spot asset whose variance follows a GARCH (p, q)-process that can be correlated with the returns of the spot asset. They found empirically for S&P500 index options that their model had lower valuation errors than the Black-Scholes-Merton model with implied volatilities. They argued that the GARCH model was able to simultaneously capture the correlation of volatility with spot returns and the path dependence in volatility. Hao and Yang (2011) presented a scenario-based risk measure for a portfolio of European-style derivative securities over a fixed time horizon under the regime-switching Black-Scholes economy. The study derived a closed-form expression for the risk measure for vanilla European options and barrier options, and this approach can be applied to some other exotic options. The results of the study provide some guidelines and insights for portfolios containing different kinds of derivatives. Jacobs and Christofferson (2004) compared a range of GARCH models with different lags, using option prices and returns. They found that, in contrast to the returns-based objective function, using an option price-based objective function favored a more parsimonious model. Jacobs et al (2004) suggested that index option prices differ systematically from those predicted by the Black-Scholes-Merton model. In particular, out-of-the-money put prices and in-the-money call prices were higher than predicted by the Black-Scholes-Merton model. They suggested an analytic option pricing formula consistent with the stock return dynamic, viz. an inverse Gaussian GARCH model, which performed better than the usual BSM model for out-of-the-money puts on the S&P 500 Index. Singh et al (2011) empirically investigated the forecasting performance of closed-form discrete time GARCH option pricing model with benchmark Black-Scholes and its version practitioner Black-Scholes model for pricing S&P Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 427 CNX Nifty 50 index option of India, relative to market price using error metrics, moneyness-maturity-wise. They found that the practitioner Black-Scholes model outperforms the other two models, and reduced the price bias between model and market. Varma (2002) evaluated the volatility pricing of the index options with the help of the Black-Scholes-Merton option pricing formula and the GARCH (1, 1) model and has found severe mispricing in Indian Index options. He has also established the significant difference in volatility smiles for call and put options. Lehar et al (2002) examined the performance of two extensions of the Black-Scholes-Merton framework, the GARCH and the stochastic volatility option pricing model. They found empirically for FTSE 100 option prices that GARCH dominated over the stochastic volatility and the Black-Scholes-Merton model. However, they found significant errors in the prediction of the market risk from hypothetical derivative positions in all the models. 3. Methodology The objective of the present study is to analyse systematic mispricing of stock and index options on the NSE using the GARCH model and the Black-Scholes-Merton options pricing model. To analyses the stock options ten companies from ten different sectors, closing stock prices were obtained from the National Stock Exchange1 for the period of 1-May-2012 to 30-Apr-2013 were taken to calculate the volatility using the GARCH(1,1) model for 30-,60-, and 90-day periods. The volatility values thus obtained were used in the Black-Scholes-Merton model to calculate the call and put prices for the stocks. 3.1. GARCH (Generalized Autoregressive Conditional Hetroscedasticity) Model The Generalized Autoregressive Conditional Heteroscedasticity (GARCH) models were propounded by Engle (1982) and Bollerslev (1986). The distinctive feature of these models is that they recognize that volatilities and correlations are not constant: i.e. volatility clustering and excess kurtosis. The GARCH models are discrete-time models, attempting to track changes in the correlation and volatility over time. The GARCH model is used to estimate volatility for a variety of financial time series: stock returns, interest rates, and foreign exchange rates. GARCH models have been applied in various fields such as asset allocation, risk management, and portfolio management, and option pricing. The GARCH (p, q) model is formulated as: , where p is the order of the GARCH (lagged volatility) terms, and q is the order of the ARCH (lagged squared-error) terms. In the academic literature, the GARCH (1, 1) process seems to be perceived as a realistic data generating process for financial returns. An intuitively appealing interpretation of the GARCH (1, l) model is easy to understand. The GARCH forecast variance is a weighted Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 428 average of three different variance forecasts. One is a constant variance that corresponds to the long-run average. The second is the forecast that was made in the previous period. The third is the new information that was not available when the previous forecast was made. This could be viewed as a variance forecast based on one period of information. The weights on these three forecasts determine how fast the variance changes with new information and how fast it reverts to its long-run mean. Volatility and risk both terms are used interchangeably today. If one decides to approach the difficult problem of forecast evaluation, the first consideration is: which volatility is being forecast? For option pricing, portfolio optimization and risk management one needs a forecast of the volatility that governs the underlying price process until some future risk horizon. Future volatility is an extremely difficult thing to forecast because the actual realization of the future process volatility will be influenced by events that happen in the future, e.g. large market movements at any time before the risk horizon. Thus the real problem is that of prediction of volatility. The predicted volatility can be used to determine future prices of the stock or the stock option, and thus an investor can use arbitrage strategies accordingly to benefit from the model. The GARCH (1, 1) model is represented as , where γ represents the weight of long run variance; VL represents the long-run variance, α the weight of periodic returns, and β the weight of variance. The parameters α, β and ω are estimated by using the Maximum Likelihood Method, maximizing the log-likelihood function , subject to the constraint α + β < 1. Once the values of α, β and ω are obtained, γ = 1 – α – β, and VL is calculated as ω/γ. The annualized volatility is calculated as 251*VL. This volatility is then used to calculate the option prices. 3.2. The Black-Scholes-Merton Model The Black Scholes-Merton model (1973) is one of the most important concepts in modern financial theory. The BSM model gives the formulae for European call and put options on a non-dividend-paying stock as follows: Where S represents the spot price of stock, X represents the exercise price of the option, r is the annual risk-free rate of return, t is the time to expiry of the option, and is the annual Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 429 volatility of the stock. In the analysis, for each option, the exercise price was taken at par with the spot price on 1-Jan-2013; the times to expiry considered were 30-, 60-, and 90-days; and the risk-free rate considered was 7.27% p.a. The volatility used was the long-run volatility estimated by the GARCH model. The market values of the options were compared with the estimated values using the paired-samples Wilcoxon test. The %age difference between the market values and the estimated GARCH-BSM prices were calculated to assess the extent of mispricing. Also, the extent of mispricing for 30-, 60-, and 90-day call and put options were compared using the paired-samples Wilcoxon test. 4. Analysis 4.1 Call Option: Table 1. Comparison of call option values calculated using Black-Scholes and the Market Option value for 30, 60 & 90 day expiry Companies Call Option Black-Scholes Option Value Market Value 30 Day 60 Day 90 Day 30 Day 60 Day 90 Day Ambuja Cements Limited 4.23 11.59 16.17 7.6 19.45 19.45 Bharti Airtel 13.1 15.51 29.2 18.65 26 38.5 Cipla Limited 12.6 20.96 28.05 17.3 35.2 35.2 DLF Limited 11.19 13.59 14.35 18.6 29 15.55 Hero MotoCorp Limited 106.18 130.69 133.8 354.25 214.45 134.4 Hindustan Petroleum Corporation Limited 24.07 22.78 17.73 23.8 38 15 Hindustan Unilever Limited 15.68 30.67 26.05 19.35 89.2 29 Infosys Limited 98.18 119.48 167.6 102.15 5 138.1 State Bank of India 52.56 63.85 100.96 92.3 150 145.3 Sun TV Network Limited 21.62 28.98 27.41 32.5 23.95 40.5 Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 430 Table 2. Average difference in three different stock call options for different time period of expiry Companies Percentage Difference 30 Day 60 Day 90 Day Ambuja Cements Limited 0.443421 0.404113 0.168638 Bharti Airtel 0.297587 0.403462 0.241558 Cipla Limited 0.271676 0.404545 0.203125 DLF Limited 0.398387 0.531379 0.07717 Hero MotoCorp Limited 0.700268 0.390581 0.004464 Hindustan Petroleum Corporation Limited -0.01134 0.400526 -0.182 Hindustan Unilever Limited 0.189664 0.656166 0.101724 Infosys Limited 0.038864 -22.896 -0.21361 State Bank of India 0.430553 0.574333 0.305162 Sun TV Network Limited 0.334769 -0.21002 0.32321 Average difference in call option prices varies based on time effect of 30, 60 & 90 days. There is only a minute difference in the option prices and the above table also shows that the stock call option with 30 days to expiry has a difference which is minimum between the model and market values. 4.2 Put Option: Table 3. Comparison of put option values calculated using Black-Scholes and the Market Option value for 30, 60 & 90 day expiry Companies Put Option Black-Scholes Option Value Market Value 30 Day 60 Day 90 Day 30 Day 60 Day 90 Day Ambuja Cements Limited 16.2 9.7 8.37 17.4 14.25 14.8 Bharti Airtel 8 10.99 17.66 16.6 25 25 Cipla Limited 8.69 8.48 21.27 15.8 19.5 30.35 DLF Limited 9.66 12.21 9.23 17 51.5 21.1 Hero MotoCorp Limited 92.63 72.79 83.57 119.6 62.05 120.95 Hindustan Petroleum Corporation Ltd. 15.75 18.76 11.29 18.85 71.05 18.5 Hindustan Unilever Limited 6.68 13.94 14.11 15.7 6.35 29.55 Infosys Limited 57.91 88.86 113.41 76.8 453.75 94.85 State Bank of India 17.44 44.99 51.38 68.15 128.7 128.7 Sun TV Network Limited 19.3 21.11 17.74 30.95 80.5 39.8 Table 4. Average difference in three different stock put options for different time period of Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 431 expiry Companies Percentage Difference 30 Day 60 Day 90 Day Ambuja Cements Limited 0.069 0.3193 0.4345 Bharti Airtel 0.5181 0.5604 0.2936 Cipla Limited 0.45 0.5651 0.2992 DLF Limited 0.4318 0.7629 0.5626 Hero MotoCorp Limited 0.2255 -0.1731 0.3091 Hindustan Petroleum Corporation Limited 0.1645 0.736 0.3897 Hindustan Unilever Limited 0.5745 -1.1953 0.5225 Infosys Limited 0.246 0.8042 -0.1957 State Bank of India 0.7441 0.6504 0.6008 Sun TV Network Limited 0.3764 0.7378 0.5543 Average difference in put option prices varies based on time effect of 30, 60 & 90 days. There is only a minute difference in the option prices and the above table also shows that the stock call option with 30 days to expiry has a difference which is minimum between the model and market values. 4.3 Paired T-Test Table 5. SPSS Output of Paired Sample T-Test to compare the model and market prices of thirty day call option price T-Test Paired Samples Statistics Mean N Std. Deviation Std. Error Mean Pair 1 ThirtydayBSM ThirtydayMV 35.9410 68.6500 10 10 37.304 50 105.602 34 11.796 72 33.394 39 Paired Samples Correlations N Correlation Sig. Pair 1 ThirtydayBSM & ThirtydayMV 10 0.849 0.002 Paired Samples Test Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 432 Paired Differences 95% Confidence interval of the Difference Pair 1 Thirtyday BSM Thirtyday MV Mean Std. Deviation Std. Error Mean Lower Upper t df Sig. (2-tailed) 32.709 00 76.511 47 24.195 05 87.442 01 22.024 01 -1.352 9 0.209 Paired sample T-test is done to check whether the numerical difference between the actual and the expected thirty day call option price of stock option which is significant in this case. The SPSS result show that the p value is greater than 0.05. So we can accept the null hypothesis that there is no significant difference between the actual and expected call option prices of stock option. Table 6. SPSS Output of Paired Sample T-Test to compare the model and market prices of sixty day call option price T-Test Paired Samples Statistics Mean N Std. Deviation Std. Error Mean Pair 1 SixtydayBSM SixtydayMV 45.8100 63.0250 10 10 44.403 36 68.232 35 14.041 57 21.576 96 Paired Samples Correlations N Correlation Sig. Pair 1 SixtydayBSM & SixtydayMV 10 0.564 0.089 Paired Samples Test Paired Differences 95% Confidence interval of the Difference Pair 1 SixtydayBSM SixtydayMV Mean Std.Deviation Std.Error Mean Lower Upper t df Sig. (2-tailed) 17.21500 56.641 61 17.911 65 57.733 96 23.303 96 -961 9 0.362 Paired sample T-test is done to check whether the numerical difference between the actual and the expected sixty day call option price of stock option which is significant in this case. The SPSS result show that the p value is greater than 0.05. So we can accept the null hypothesis that there is no significant difference between the actual and expected call option Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 433 prices of stock option. Table 7. SPSS Output of Paired Sample T-Test to compare the model and market prices of ninety day call option price T-Test Paired Samples Statistics Mean N Std. Deviation Std. Error Mean Pair 1 NinetydayBSM NinetydayMV 56.1320 61.1000 10 10 56.300 98 54.734 31 17.803 93 17.308 51 Paired Samples Correlations N Correlation Sig. Pair 1 NinetydayBSM & NinetydayMV 10 0.948 0.000 Paired Samples Test Paired Differences 95% Confidence interval of the Difference Pai r 1 NinetydayBS M NinetydayM V Mean Std.Deviati on Std.Error Mean Lower Upper t d f Sig. (2-tailed) -4.968 00 18.034 38 5.702 97 -17.869 02 7.933 02 -87 1 9 0.406 Paired sample T-test is done to check whether the numerical difference between the actual and the expected ninety day call option price of stock option which is significant in this case. The SPSS result show that the p value is greater than 0.05. So we can accept the null hypothesis that there is no significant difference between the actual and expected call option prices of stock option. Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 434 Table 8. SPSS Output of Paired Sample T-Test to compare the model and market prices of thirty day put option price T-Test Paired Samples Statistics Mean N Std. Deviation Std. Error Mean Pair 1 ThirtydayBSM ThirtydayMV 25.2260 39.6850 10 10 27.958 25 36.170 13 8.841 18 11.438 00 Paired Samples Correlations N Correlation Sig. Pair 1 ThirtydayBSM & ThirtydayMV 10 0.925 0.000 Paired Samples Test Paired Differences 95% Confidence interval of the Difference Pair 1 ThirtydayBSM ThirtydayMV Mean Std.Deviation Std.Error Mean Lower Upper t df Sig. (2-tailed) -14.459 00 14.800 94 4.680 47 -25.046 96 -3.871 04 -3.089 9 0.013 Paired sample T-test is done to check whether the numerical difference between the actual and the expected thirty day put option price of stock option which is significant in this case. The SPSS result shows that the p value is less than 0.05. So we can reject the null hypothesis that there is a significant difference between the actual and expected call option prices of stock option. Table 9. SPSS Output of Paired Sample T-Test to compare the model and market prices of sixty day put option price T-Test Paired Samples Statistics Mean N Std. Deviation Std. Error Mean Pair 1 SixtydayBSM SixtydayMV 30.1830 91.2650 10 10 28.937 11 132.681 40 9.150 72 41.957 54 Paired Samples Correlations N Correlation Sig. Pair 1 SixtydayBSM & SixtydayMV 10 0.793 0.006 Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 435 Paired Samples Test Paired Differences 95% Confidence interval of the Difference Pair 1 SixtydayBSM SixtydayMV Mean Std.Deviation Std.Error Mean Lower Upper t df Sig. (2-tailed) 61.082 00 111.134 79 35.143 91 -140.583 18.419 04 -1.738 9 0.116 Paired sample T-test is done to check whether the numerical difference between the actual and the expected sixty day put option price of stock option which is significant in this case. The SPSS result show that the p value is greater than 0.05. So we can accept the null hypothesis that there is no significant difference between the actual and expected put option prices of stock option. Table 10. SPSS Output of Paired Sample T-Test to compare the model and market prices of ninety day put option price T-Test Paired Samples Statistics Mean N Std. Deviation Std. Error Mean Pair 1 NinetydayBSM NinetydayMV 34.8030 52.3600 10 10 36.414 27 44.451 33 11.515 20 14.056 74 Paired Samples Correlations N Correlation Sig. Pair 1 NinetydayBSM & NinetydayMV 10 0.823 0.003 Paired Samples Test Paired Differences 95% Confidence interval of the Difference Pair 1 NinetydayBSM NinetydayMV Mean Std.Deviation Std.Error Mean Lower Upper t df Sig. (2-tailed) -17.557 00 25.239 35 7.981 38 -35.61214 498 14 -2.200 9 0.55 Paired sample T-test is done to check whether the numerical difference between the actual and the expected ninety day put option price of stock option which is significant in this case. The SPSS result show that the p value is greater than 0.05. So we can accept the null Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 436 hypothesis that there is no significant difference between the actual and expected put option prices of stock option. 4.4. MULTIPLE REGRESSIONS Table 11. SPSS output of Multiple Regression for Stock Call Options: Regression Variables Entered/ Removed ᵇ Model Variables Entered Variables Removed Method 1 Maturity, Stock, Price, Volatility, Strike Price ᵃ Enter a. All reserved variables entered b. Dependent Variable: Option Price Model Summary Model R R Square Adjusted R square Std. Error of the estimate 1 0.971ᵃ 0.943 0.934 11.73665 a. Predictors: ( Constant), maturity, Stock Price, Volatility, Strike Price ANOVAᵇ Model Sum of Squares df Mean Square F Sig. 1 Regression 57392.765 4 14348.191 104.162 0.000ᵃ Residual 3443.722 25 137.749 Total 60836.487 29 a. Predictors: ( Constant ), maturity, stock Price, Volatility, Strike Price b. Dependent Variable: Option Price Multiple regression is done to find out the independent variables on which call option prices of the stock option depends upon. The independent variables considered are strike price, spot price, volatility and maturity time. The results of the SPSS output show there is dependency of call option prices of the stock option on all the variables except the maturity time as R square value is high and the p values are less than 0.05 in all the cases except maturity time. Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 437 Table 12. SPSS output of Multiple Regression for Stock Put Options: Regression Variables Entered/ Removed ᵇ Model Variables Entered Variables Removed Method 1 Maturity, Stock, Price, Volatility, Strike Price ᵃ Enter a. All reserved variables entered b. Dependent Variable: Option Price Model Summary Model R R Square Adjusted R square Std. Error of the estimate 1 0.966ᵃ 0.932 0.922 8.5351 a. Predictors: ( Constant), maturity, Stock Price, Volatility, Strike Price ANOVAᵇ Model Sum of Squares df Mean Square F Sig. 1 Regression 25142.779 4 14348.191 86.286 0.000ᵃ Residual 1821.177 25 72.847 Total 26963.957 29 a. Predictors: ( Constant ), maturity, stock Price, Volatility, Strike Price b. Dependent Variable: Option Price Coefficientsᵃ Unstandardized Coefficients Standardized Coefficients Model 1 ( Constant) B Std. Error Beta t Sig. -28.825 5.624 -5.125 0 Stock Price -0.742 0.08 -22.412 -9.277 0 Strike price 0.755 0.079 23.187 9.611 0 Volatility 307.237 54.425 0.436 5.645 0 Maturity -72.704 32.503 -0.163 -2.237 0.034 Multiple regression is done to find out the independent variables on which call option prices of the stock option depends upon. The independent variables considered are strike price, spot price, volatility and maturity time. The results of the SPSS output show there is dependency of put option prices of the stock option on all the variables as R square value is high and the p Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 438 values are less than 0.05 in all the cases. 5. Discussion The findings of the study suggest that options are significantly overpriced. However, an interesting possibility suggested by the findings is that this overpricing decreases with expiration period. Also, the findings suggest that put overpricing is significantly higher than call overpricing, as suggested by Dash et al (2012), particularly for longer expiration periods. The study has several limitations. The sample size used for the analysis is small, and the selected stocks are all large-cap stocks; so that it is not clear whether the results of the study extend to medium- and small-cap stocks. Another difficulty is that of trading volume, which may also affect overpricing, as suggested by Dash et al (2012). Finally, another limitation that may bias the results of the study is the choice of research period; it is not clear whether the results extend to other periods, particularly under high volatility. There is great scope for applying GARCH option pricing models to examine several other interesting properties of options, including implied volatility, volatility smiles, and the time-variability of options properties (e.g. Greeks). References Adesi G.B, Engle. R.F And Mancini.L. (2008). A GARCH option pricing model with filtered historical simulation. Review of financial Studies. Bakshi.G, C. Cao, & Z. Chen. (1997). Empirical performance of alternative option pricing models. Journal of Finance, 52, 2003-2049. http://dx.doi.org/10.1111/j.1540-6261.1997.tb02749.x Black.F., & M.Scholes. ( 1973). The valuation of option and Corporate Liabilities, Journal of Political Economy, 81,637-654. http://dx.doi.org/10.1086/260062 Christoffwerson.P., & K. Jacobs. (2004). Which GARCH Model for option Valuation. Journal of Management Science, 50,1204-1221. http://dx.doi.org/10.1287/mnsc.1040.0276 Dash.M, Dagha.J.K., Sharma.P., & Singhal.R. (2012). An application of Garch models in detecting systematic bias in option pricing and determining arbitage in options. Journal of CENTRUM CATHEDRA, 5(1),91-101. http://dx.doi.org/10.7835/jcc-berj-2012-0069 Duan.J. (1995). The GARCH option Pricing Model. Mathematical Finance, 5(1),13-32. E.Ghysels, & Chernov.M. (2000). A study towards a unified approach to the joint estimation of objective and risk neutral measues for the purpose of option valuation. Journal of Financial Economic, 56,407-458. http://dx.doi.org/10.1016/S0304-405X(00)00046-5 Hao. F., & Yang, H. (2011). Coherent risk measure for derivatives under Black-Scholes economy with regime switching. Journal of Managerial Finance, 37(11),1011-1024. http://dx.doi.org/10.1108/03074351111167910 Heston.S.L. (1993). A close form solution for option with stochastic volatility with Asian Journal of Finance & Accounting ISSN 1946-052X 2014, Vol. 6, No. 1 www.macrothink.org/ajfa 439 application to bond and currency options. Review of financial studies, 6, 327-343. http://dx.doi.org/10.1093/rfs/6.2.327 Jackwerth, & Buraschi. A. (2001). The Price of a smile: Hedgeing and spanning in option market, 14,495-527. Jacobs.C., & Christofferson.P. (2004). Which GARCH Model for option valuation. Jounal of management sciences, 50,1204-1221. http://dx.doi.org/10.1287/mnsc.1040.0276 Jacobs.C, Heston, S.L., & Christofferson. P.. a. (2004). A GARCH Option Pricing Model with filtered historical simulation. http://dx.doi.org/10.1287/mnsc.1040.0276 Lehar.A, Scheicher.M., & Schittenkopt.C. (2002). GARCH vs. Stochastic Volatility- Option Pricing and risk management. Journal of Banking & Finance, 26(2/3), 323-346. http://dx.doi.org/10.1016/S0378-4266(01)00225-4 M.Scholes, & Black.F. (1973). Valuation of options and Corporate liabilities. Journal of political economy, 81,637-654. http://dx.doi.org/10.1086/260062 Nandi.S., & Heston. S. L. (2000). A close form GARCH Option Valuation Model. Review of financial studies, 13(3). http://dx.doi.org/10.1093/rfs/13.3.585 Ross.S.A. Cox. J.C. (1976). The valuation of options for alternative stochastic processes. Journal of Financial Economics, 3, 145-166. http://dx.doi.org/10.1016/0304-405X(76)90023-4 Singh.V.P, Ahmed.N, and Pachori.P.(2011). Empirical analysis of GARCH and practitioner Black-Scholes Model for pricing S& P Nifty 50 index options of India. Decision, 38(2),51-67. Siu.T.K. Tong. H and Yang.H. (2004). On pricing derivatives under GARCH Models: A dynamic Gerber-Shiu's Approach. North American Actuarial journal, 8, 17-31. Varma.J.R. (2002). Mispricing of Volatility in the Indian Index Option Market. IIM, Ahmedabad, Working Paper Series No: 2002-04-01.