Error diagnostic for weighted moving average to forecast Shariacompliant securities in Malaysian Stock Exchange

Size: px

Start display at page:

Download "Error diagnostic for weighted moving average to forecast Shariacompliant securities in Malaysian Stock Exchange"

Johnathan Stanley
5 years ago
Views:

Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016 Page 29-37 Error diagostic for weighted movig average to forecast Shariacompliat securities i Malaysia Stock Exchage Nashirah

1 Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016 Page Error diagostic for weighted movig average to forecast Shariacompliat securities i Malaysia Stock Exchage Nashirah Biti Abu Bakar 1 ad Sofia Rosbi 2 1 Islamic Busiess School, College of Busiess, Uiversiti Utara Malaysia, Kedah, Malaysia. 2 Uiversiti Malaysia Perlis, Perlis, Malaysia. Correspodig ashirah@uum.edu.my Article Iformatio Keywords Forecastig, Movig average, Share price, Sharia-compliat securities, Malaysia Stock Exchage. Abstract This paper forecast the share price for Sharia-compliat compaies that issues Iitial Public Offerigs (IPO) i year This paper usig forecastig methods which are simple movig average ad weighted movig average. The, these share prices are evaluated usig three types of error, which are mea absolute deviatio (MAD), mea squared error (MSE) ad mea absolute percetage error (MAPE). The results show weighted movig average is more accurate tha simple movig average. This result is to help ivestors to determie ad select share price with higher retur usig a accurate method. INTRODUCTION Islamic capital market is oe of the istrumets i Islamic ivestmet. Islamic capital market is defied as the activities cocered with the public offerig ad tradig of securities with is i lie with sharia law. Islamic capital market also defied as the market for a variety of log-term Islamic fiacial istrumets that ca be traded, either i the form of sukuk (Islamic bod) or Islamic equity capital (Astuty, 2015). Islamic capital market must sigify with sharia law where the capital must be free from ay prohibited elemet such as usury (riba), gamblig (maisir), ad ucertaities (gharar) (Abdul Rahim & Yog, 2010). Iitial Public Offerigs (IPO) for sharia-compliat compaies is part of Islamic equity capital. IPO for shariacompliat compaies is act as a alterative source for Islamic compaies to raise a capital for expase a busiess. IPO for sharia-compliat compaies was issued by compaies to the public ivestors. This activities was ivolved several legal istitutioal such as Miistry of Iteratioal Trade ad Idustry (MITI), uderwriter istitutio ad Malaysia Stock Exchage. A large umber of studies have examied the performace of share price of IPO (Boulto, et al., 2010). This iterest may be related to the importace of IPO market for ecoomic growth ad employmet, but ofte the focus is o the tremedous profit opportuities that IPO frequetly offer to ivestors (Bessler & Thies, 2007). I the same, the studies that focus o the performace of log-term IPO i Malaysia foud evidece that log-term performaces of IPO are performed better (How, et al., 2007). However, the area of forecastig for share price of sharia-compliat i Malaysia is still ot eough. Thus, this study tries to fulfill this gap by forecast the log-term performace of IPO for sharia-compliat compaies usig weighted movig average. Specifically, this study examie log term performace of IPO for 29

2 Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016, Page sharia-compliat compaies usig movig average methods ad compare with weighted movig average. Next, this study performed the aalysis of the error i the aalysis. LITERATURE REVIEW Various studies ivestigated the log term performace of IPO market foud IPO are uderperformed their bechmark. Ritter (1991) examies log term performace of IPO i Uited States foud the sigificat uderperform their bechmark. Carter et al. (1998) also foud IPO stocks are uderperformed relative to the market over a three-year holdig period is less severe for IPO hadled by more prestige uderwriters. The same result is reported by Cha, et al. (2004) i Chia. They foud IPO i Chia is slightly uderperformed. While, Lee, et al. (1996) report that Australia s IPO sigificatly uderperform market movemets i the three-year period subsequet to list. Goerge, et al. (2007) discusses the log-ru performace of UK IPO ad foud that the percetage of equity issued ad the degree of multiatioality of a firm are the key predictors of IPO performace i UK. Drobetz, et al. (2005) foud log ru uderperformace to the fact that IPO firms ted to be small firms whe used a small capitalizatio idex as a bechmark.. Thus, uderperformed of log term IPO are observed i various coutries (Kirkulak & Davis, 2005). Therefore, it is importat to forecast the performace of share price to make sure that particular share ca give high retur. Stock markets are iflueced by complex factors ad oliear relatioships amog factors existed i differet periods such that forecastig the future stock treds for ivestors are extremely difficult (Kluger & McBride, 2011). I additio, implemetatio of accurate forecastig methods is importat i predictig stock prices sice the stock market begis (Pavlov & Hur, 2012). I this paper, we compare the simple movig average ad weighted movig average. The, three types of error are calculated. The three types of error are mea absolute deviatio (MAD), mea squared error (MSE) ad mea absolute percetage error (MAPE). METHODOLOGY I Malaysia Stock Exchage, there are 13 sharia-compliat compaies that issues Iitial Public Offerig i year of Table 1 show the list of the compaies. The, for every sector, oe represetative is selected for the aalysis. As show i Table 1, Setoria Group Compay is selected for property sector, Globaltec Formatio Berhad is selected for idustrial products sector, Felda Global Veture Holdig Berhad is selected for platatio sector, OCK Group Berhad is selected for tradig/ service sector ad Gabuga AQRS Berhad is selected for costructio sector. TABLE I : List of Sharia compliace compaies that issues Iitial Public Offerigs (IPO) i year of 2012 No. Stock Code Compay Board Sector SENTORIA GROUP BERHAD MAIN MARKET PROPERTY EITA RESOURCES BERHAD MAIN MARKET TRADING/SERVICES SAPURAKENCANA PETROLEUM BHD MAIN MARKET TRADING/SERVICES PESTECH INTERNATIONAL BERHAD MAIN MARKET TRADING/SERVICES GLOBALTEC FORMATION BERHAD MAIN MARKET INDUSTRIAL PRODUCTS GAS MALAYSIA BERHAD MAIN MARKET TRADING/SERVICES FELDA GLOBAL VENTURES HLDG BHD MAIN MARKET PLANTATION OCK GROUP BERHAD ACE MARKET TRADING/SERVICES IHH HEALTHCARE BERHAD MAIN MARKET TRADING/SERVICES GABUNGAN AQRS BERHAD MAIN MARKET CONSTRUCTION PASUKHAS GROUP BERHAD ACE MARKET TRADING/SERVICES DATASONIC GROUP BERHAD MAIN MARKET TRADING/SERVICES HIAP HUAT HOLDINGS BHD ACE MARKET INDUSTRIAL PRODUCTS The, this research aalyse the movemet of share price, startig from the first moth util more tha 50 moths. The data are collected from Datastream (Thomso Reuters). The, this research calculates two methods of forecastig. The first method is simple movig average ad the secod method is weighted movig average. The calculatio for simple movig average is as follows: demad i previous periods Movig average = (1) 30

3 Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016, Page where is the umber of periods i the movig average. I this research, =3 moths is selected. Next, the calculatio for weighted movig average is as follows: Weighted movig average = (Weight for period )(Demad i period ) Weights (2) where is the umber of periods i the movig average. I this research, =3 moths is selected. Weight for period 1, period 2 ad period 3 is 0.05, 0.15 ad 0. This settig is to make this forecastig method more resposive to chages because more recet periods are set to be more heavily weighted. The, to validate the robustess diagostic of forecastig method, overall accuracy eed to be determied. The forecast error ca be determied by comparig the forecasted values with the actual or observed values. If F t deotes the forecast i period t, ad A t deotes the actual value i period t, the forecast error is defied as: Forecast error = Actual value - Forecast value A F t t (3) The, to validate the robustess diagostic of forecastig method, three types of error is itroduced. The first measure of the overall forecast error for a model is the mea absolute deviatio (MAD). This value is calculated by takig sum of the absolute values of the idividual forecast errors ad dividig by the umber of periods of data (): Forecast error MAD = = t1 t1 Actual-Forecast At Ft t 1 Next, the secod parameter for measurig overall forecast error is mea squared error (MSE). The MSE is average of squared differeces betwee the forecasted ad observed values. The formula for MSE is: (4) MSE = Forecast error Actual - Forecast t1 t1 t 1 A F t t (5) The, the third parameter for measurig forecast error is mea absolute percetage error (MAPE). The MAPE is computed as the average of the absolute differece betwee the forecasted ad actual values, expressed as a percetage of the actual values. The MAPE is calculated usig ext equatio: 100 Forecast error / Actual 100 Actual Forecast / Actual MAPE = = t t t t t1 t1 t A F / A t t t (6) As robustess for diagostic error checkig, three method of error calculatio is compared to fid the lowest error for forecastig methods. I this paper, the accuracy betwee simple movig average ad weighted movig average are evaluated. 31

4 Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016, Page RESULT AND DISCUSSION Property Sector I the robust diagostic aalysis of errors, oe compay that represet the performace of property sector is selected. The compay that shows the average performace of property sector is Setoria Group Berhad Movig average Weighted movig average Period (moth) Period (moth) -0.3 Movig average Weighted movig average Error evaluatio Movig average Weighted movig average Mea absolute deviatio (MAD) Mea squared error (MSE) Mea absolute percetage error (MAPE) Figure 1: Share price for Setoria Group Berhad Real dyamic movemet of share price Forecast error aalysis Evaluatio for three types of error Figure 1 shows the movemet of real share price for Setoria Group Berhad. The maximum value is MYR 1.54 i September 2014 (32th moth). The miimum value is MYR i Jue 2012 (5th moth). Figure 1 shows the forecast error from fourth moth to 57th moth. The result shows the forecast error for weighted movig average is lower tha movig average. Figure 1 is the table that idicated the three types of errors. Result shows three types of errors i weighted movig average method are lower tha movig average method. Therefore weighted movig average is more suitable for forecastig share price of compay i property sector. Idustrial product sector 32

5 Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016, Page I the robust diagostic aalysis of errors, oe compay that represet the performace of idustrial product sector is selected. The compay that shows the average performace of idustrial product sector is Globaltec Formatio Berhad Movig average Weighted movig average 0.04 Period (moth) Movig average Weighted movig average Period (moth) Error evaluatio Movig average method Weighted movig average Mea absolute deviatio (MAD) Mea squared error (MSE) Mea absolute percetage error (MAPE) Figure 2: Share price for Globaltec Formatio Berhad Real dyamic movemet of share price Forecast error aalysis Evaluatio for three types of error Figure 2 shows the movemet of real share price for Globaltec Formatio Berhad. The maximum value is MYR i May 2012 (first moth). The miimum value is MYR i September 2016 (53th moth). Figure 2 shows the forecast error from fourth moth to 54th moth. The result shows the forecast error for weighted movig average is lower tha movig average. Figure 2 is the table that idicated the three types of errors. The MAD ad MAPE of weighted movig average method are less tha movig average method. Therefore weighted movig average is more suitable for forecastig share price of compay i idustrial product sector. Platatio sector 33

6 Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016, Page I the robust diagostic aalysis of errors, oe compay that represet the performace of platatio sector is selected. The compay that shows the average performace of platatio sector is Felda Global Vetures Holdig Berhad. 6 5 Movig average Weighted movig average Period (moth) Period (moth) Movig average Weighted movig average Error evaluatio Movig average method Weighted movig average Mea absolute deviatio (MAD) Mea squared error (MSE) Mea absolute percetage error (MAPE) Figure 3: Share price for Felda Global Vetures Holdig Berhad Real dyamic movemet of share price Forecast error aalysis Evaluatio for three types of error Figure 3 shows the movemet of real share price for Felda Global Vetures Holdig Berhad. The maximum value is MYR i July 2012 (Secod moth). The miimum value is MYR i May 2016 (48th moth). Figure 3 shows the forecast error from fourth moth to 53th moth. The result shows the forecast error for weighted movig average is lower tha movig average. Figure 3 is the table that idicated the three types of errors. Result shows three types of errors i weighted movig average method are lower tha movig average method. Therefore weighted movig average is more suitable for forecastig share price of compay i platatio sector. 34

7 Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016, Page Tradig/service sector I the robust diagostic aalysis of errors, oe compay that represet the performace of tradig/service sector is selected. The compay that shows the average performace of tradig/service sector is OCK Group Berhad Movig average Weighted movig average Period (moth) 0.3 Movig average Weighted movig average Period (moth) - Error evaluatio Movig average method Weighted movig average Mea absolute deviatio (MAD) Mea squared error (MSE) Mea absolute percetage error (MAPE) Figure 4: Share price for OCK Group Berhad Real dyamic movemet of share price Forecast error aalysis Evaluatio for three types of error Figure 4 shows the movemet of real share price for OCK Group Berhad. The maximum value is MYR 66 i July 2014 (25th moth). The miimum value is MYR 47 i July 2012 (first moth). Figure 4 shows the forecast error from fourth moth to 52th moth. The result shows the forecast error for weighted movig average is lower tha movig average. Figure 4 is the table that idicated the three types of errors. Result shows three types of errors i weighted movig average method are lower tha movig average method. Therefore weighted movig average is more suitable for forecastig share price of compay i tradig/service sector. 35

8 Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016, Page Costructio sector I the robust diagostic aalysis of errors, oe compay that represet the performace of costructio sector is selected. The compay that shows the average performace of costructio sector is Gabuga AQRS Berhad Movig average Weighted movig average Period (moth) Period (moth) Movig average Weighted movig average Error evaluatio Movig average method Weighted movig average Mea absolute deviatio (MAD) Mea squared error (MSE) Mea absolute percetage error (MAPE) Figure 5: Share price for Gabuga AQRS Berhad Real dyamic movemet of share price Forecast error aalysis Evaluatio for three types of error Figure 5 shows the movemet of real share price for Gabuga AQRS Berhad. The maximum value is MYR i September 2014 (27th moth). The miimum value is MYR 35 i Jauary 2016 (43rd moth). Figure 5 shows the forecast error from fourth moth to 52th moth. The result shows the forecast error for weighted movig average is lower tha movig average. Figure 5 is the table that idicated the three types of errors. Result shows three types of errors i weighted movig average method are lower tha movig average method. Therefore weighted movig average is more suitable for forecastig share price of compay i costructio sector. 36

9 Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016, Page CONCLUSION The objective of this research is to evaluate the forecastig method to predict the log term performace of share price. The share prices that ivolved i this research are sharia compliace compaies that listed i Malaysia Stock Exchage i year of I this research, we aalyzed the implemetatio of two methods of forecastig which are simple movig average ad weighted movig average. From the result i Figure 1 util Figure 5, this research cocludes that: (1) The value of share price is dyamically movig fluctuate accordig to market coditio. (2) The evaluatio of each forecastig method is evaluate thoroughly with respect to three measure of error calculatio, which are mea absolute deviatio (MAD), mea squared error (MSE) ad mea absolute percetage error (MAPE). (3) The error value for each calculatio is bigger if the volatility of error for the share price is higher. (4) The forecastig method of weighted movig average is better tha simple movig average because of it is less error. Therefore, weighted movig average is more accurate i forecastig the dyamic behavior of share price. REFERENCES Abdul Rahim, R., & Yog, O. (2010). Iitial returs of Malaysia IPOs ad Shari a-compliat status. Joural of Islamic Accoutig ad Busiess Research, 1(1), Abu Bakar, N. ad Rosbi, S. (2016). Log Term Performace of Islamic Share Price for Iitial Public Offerigs (IPOs) i Malaysia: Evidece from Sharia-Compliat Compaies Listed o the Malaysia Stock Exchage ( ). Iteratioal Joural of Maagemet Sciece ad Busiess Admiistratio, 2(4), Ahmad Zaluki, N. A, Campbell, K. ad Goodacre, A. (2007). The Log Ru Share Price Performace of Malaysia Iitial Public Offerigs (IPOs). Joural of Busiess ad Accoutig, 34(1), Astuty, W. (2015). The Extraordiary Solutio for Idoesia Ecoomic Crisis: Shariah Capital Market. Joural of Fiace ad Bak Maagemet, 3(2), Boulto, T. J., Smart, S. B., & Zutter, C. J. (2010). IPO Uderpricig ad Iteratioal Corporate Goverace. Joural of Iteratioal Busiess Studies, 41(2), Bessler, W. ad Thies (2007). The log-ru performace of iitial public offerigs i Germay. Maagerial Fiace, 33(6), Carter, R.B., Dark F. H., ad Sigh, A.K. (1998). Uderwriter Reputatio, Iitial Retur ad Log Ru Performace of IPO stocks. The Joural of Fiace, 53(1), Cha, K., Wag, J. ad Wei K.C.J. (2004). Uderpricig ad log term performace of IPOs i Chia. Joural of Applied Corporate Fiace, 10(3), Che, A. Che, L.W. ad Kao, L (2010). Leverage, liquidity ad IPO log ru performace: evidece from Taiwa IPO markets. Iteratioal Joural of Accoutig & Iformatio Maagemet, 18(1), Corhay, A., Teo, S. ad Rad, A.T. (2002). The log ru performace of Malaysia iitial public offerigs (IPO): value ad growth effects. Maagerial Fiace, 28(2), Drobetz, W., Kammerma, M. ad Walchle, U. (2005). Lo ru performace of Iitial Public Offerigs: The evidece for Switzerlad. Schmalebach Busiess Review, 57, Goerge, M. Khurshed, A ad Mudambi, R. (2007). The log ru performace of UK: ca it be predicted. Maagerial Fiace, 33(6), How, J., Jelic, R., Saadoui, B., & Verhoeve, P. (2007). Share allocatios ad performace of KLSE secod board IPOs. Pacific-Basi Fiace Joural, 15(3), Lee, P.J, Taylor, S.L. ad Walter, T.S. (1996). Australia IPO pricig i the short ad log ru, Joural of Bakig ad Fiace, 20, Kluger, B.D ad McBride, M, E. (2011). Itraday tradig patters i a itelliget autoomous aget-based stock market, Joural of Ecoomic Behavior & Orgaizatio, 79, Kirkulak, B., ad Davis, C. (2005). Uderwriter Reputatio ad Uderpricig: Evidece from the Japaese IPO Market. Pacific-Basi Fiace Joural, 13, Kim, J., Krisky, I., ad Lee, J. (1995). The aftermarket performace of iitial public offerigs i Korea, Pacific Basi Fiace Joural, 3, Pavlov, V ad Hur, S. (2012). Testig the profitability of movig-average rules as a portfolio selectio strategy, Pacific-Basi Fiace Joural, 20, Paudyal, K., Saadoui. B. ad Bristo, R.J. (1998). Privatisatio Iitial Public Offerig i Malaysia: Iitial Premium ad Log Term Performace, Pacific-Basi Fiace Joural, Vol. 6, No. 5, pp Ritter, J.R. (1991). The Log Ru Performace of Iitial Public offerigs, Joural of Fiace, 46(1), 3 27 Welch, I ad Ritter, J. (2002). A Review of IPO Activity, Pricig ad Allocatios, Yale ICF Workig Paper, No

Reliability of Exponential Smoothing Method for forecasting Islamic Share Price to oil and gas sector in Malaysian Stock Exchange

Reliability of Exponential Smoothing Method for forecasting Islamic Share Price to oil and gas sector in Malaysian Stock Exchange Iteratioal Academic Research Joural of Busiess ad Techology 2(2) 2016 Page 38-44 Reliability of Expoetial Smoothig Method for forecastig Islamic Share Price to oil ad gas sector i Malaysia Stock Exchage