Pengujian anomali winner-loser pada portofolio saham berbasis Jakarta Islamic Index dan Non Jakarta Islamic Index di Bursa Efek Indonesia

KUSUMA, Taufik Hendra, Hardo Basuki, Dr., M.Soc.Sc

2010 | Tesis | S2 Magister Manajemen

Penelitian ini bertujuan untuk mengetahui ada tidaknya anomaJi winnerloser berupa pembalikan average abnormal return (AAR) pada portofolio saham berbasis Jakarta Islamic Index (JII) dan non JII. Pemilihan portofolio berbasis saham JII disebabkan oleh asumsi awal bahwa sikap hati-hati dalam pemilihan instrumen saham Jll menimbulkan dugaan bahwa tingkat return aham-saham yang terdapat daJam JII cenderung lebih stabil dan cenderung tidak memiliki efek pembaJikan harga yang signifikan dibandingkan portofolio saham lainnya. Metode penelitian yang digunakan dalam penelitian ini adalah metode deskriptif analitis ex-facto , yaitu pengumpulan data di lakukan setelah semua kejadian selesai berlangsung. Model penentuan abnormal return yang digunakan adaJah model penyesuaian pa ar (market adjusted model). Data perdagangan yang digunakan adalah data harga saham individual harian untuk menghitung return individual saham dan data Indkes Harga Saham Gabungan (IHSG) untuk menghitung return pasar untuk periode J Januari 2004 sampai dengan 30 Juni 2010. Sampel awal berupa 28 saham berbasis JII dan 28 saham berba is non JII dipilih dengan metode purposive sampling. Sampel saham berbasi JII dan non HI tersebut dibentuk menjadi 3 jenis portofolio yang terdiri dari 4, 6 dan 8 saham berbasis JII dan non JII sesuai kriteria saham winner dan loser dengan holding period pengujian selama 1, 2 dan 3 bulan. Jumlah portofolio untuk holding period I, 2 dan 3 bulan masing-masing adalah 77, 38 dan 25 portofolio saham winner dan loser berbasis JII dan non Jll. Has il pengujian hipotesis atas ada tidaknya pembalikan harga saham winner dan loser menunjukkan adanya anomali winner-loser pada portofolio saham berbasis JII dan non JII untuk holding period I, 2 dan 3 bulan. Hasil pengujian atas ada tidaknya reak i berlebihan menunjukkan kondisi aham loser mengungguli saham winner pada portofolio saham berbasis non Jll namun sebal iknya tidak terjadi pad a portofolio berbasis J II. Pad a saham berbasis JII pembalikan AAR portofolio saham winner lebih besar dari pembalikan AAR portofolio saham loser. Pada saham berbasis non Jll pembalikan AAR portofolio saham winner lebih kecil dari pembalikan AAR portofolio saham loser. Hasil uji t sampel berpasangan pada hipotesis perbedaan abnormal return saham winnerloser pada periode formasi dan pengujian menunjukkan bahwa secara statistik terdapat perbedaan yang signifikan antara rata-rata AAR periode formasi dan pengujian. Hal ini menguatkan adanya fenomena pembalikan rata-rata AAR portofolio pada periode formasi dan pengujian. Hasil uji t sampel independen atas hipotes is perbedaan abnormal return saham winner dan loser pada periode penguji an menunjukkan bahwa tidak terdapat perbedaan signifikan antara abnormal return pada portofolio berbasis Jll dan sebaliknya terdapat perbedaan signifikan pada portofolio aham berbas is non JIJ. Untuk porto fo lio saham berbasis Jll, hal ini menguatkan hasil pengujian ebelumnya, dimana reaksi berlebihan terjadi pada saham winner dan loser. Ha il uj i t sampel independen atas hipotesis perbedaan abnormal return saham berbasis JII dan non Jll pada periode pengujian menunjukkan bahwa tidak terdapat perbedaan s ignifikan atas rata-rata abnormal return portofolio saham loser berbasis JII dan non JII. Kondisi berbeda terjadi pada portofolio saham winner dimana pada holding period I dan 2 bulan terjadi perbedaan signifikan rata- rata abnormal return masing-masi ng pada portofolio 6 dan 8 saham serta 4 dan 6 saham. Untuk holding period 3 bulan, perbedaan signifikan rata-rata abnormal return terjadi pada seluruh portofolio saham loser. Simpulan yang diperoleh dari penelitian ini adalah bahwa asumsi portofol io saham berbasis JII cenderung lebih tahan terhadap reaksi berlebihan cenderung tidak terbukti, terutama untuk saham loser. Hasi l uji statistik pada hipotesis 3 menunjukkan bahwa terdapat perbedaan signi fikan atas rata-rata AAR periode formasi dan pengujian untuk seluruh portofolio berbasis JII dan non Jll. Hasil uji statistik pada hipotesis 5 menunjukkan secara lebih spesifik bahwa ke amaan adanya reaksi berlebihan portofolio berbasis JII dan non JII tersebut terjadi pada saham Loser, ditandai dengan tidak adanya perbedaaan signifikan secara statistik atas rata-rata AAR saham loser berbas is J II dan non JII. Kondi i ini juga mengindikasikan bahwa saham-saham berba i non JI1 cenderung terbia a dengan informasi bad news dan sens itif terhadap informasi good news, terutama apabila saham-saham berbasis non JII tersebut sedang berada dalam kondisi loser. Selain itu diperoleh data bahwa dengan semakin banyaknya jumlah saham dalam portofolio maka semakin sering terjadi kondi si tidak terjadi pembalikan. Hal ini menguatkan pentingnya strategi diversifikasi portofolio untuk hasil in vestasi yang lebih stabil. Khusus untuk portofo lio saham winner, semakin lama holding period dalam portofolio maka perbedaan rata-rata abnormal return antara periode formasi dan pengujian akan semakin besar dengan kemungkinan abnormal return terbesar akan dapat diperoleh atas investasi pada saham-saham dengan holding period 3 bulan. Untuk portofolio aham loser tidak terdapat kecenderungan seperti saham winner. Kata kunci: anomali winner-Loser, abnormal reum, portofolio, Jakarta Islamic Index, JII, holding period, good news, bad news, pembalikan, reaksi berlebihan.

This research aims at knowing the existence of the winner-loser anomaly in the form of reversal average abnormal return (AAR) to share portfolio based on Jakarta islamic Index (Jll) and non Jll. Choosing portfolio based on Jll share is caused by early supposition that careful act in choosing J/1 share caused an assumption that return rate of the shares in the J/1 is more stable and has no significant price reversal effect compared to another share portfolio. This research uses an ex-facto analytical descriptive method that is data collection is done upon events completed. Abnormal return determination model uses adjusted market model method. The trade data used is daily individual share price data to count share individual return and Jakarta Composite Index (IHSG) data to count market return for the period between January I, 2004 and June 30, 2010. Initial sample in the form of 28 shares based on J/1 and 28 shares based on non J/1 are chosen using purposive sampling method. Such sample of share based on J/1 and non Jll is formed to become 3 portfolio types which consist of 4, 6 and 8 shares based on Jll and non Jll in accordance with criteria of the winner and loser share with holding period test for I ,2 and 3 months. The number of portfolio for holding period of 1,2 and 3 months is respectively 77, 38 and 25 portofolio of the winner dan loser share based on Jll and non Jll. Result of hypothesis test to the existence of reversal AAR of the winner and loser share shows the existence of winner-loser anomaly to portfolio of share based on Jll and non Jll for holding period/, 2 and 3 months. Result of test to the existence of overreaction shows the loser share is higher than the winner share in portfolio of share based on non Jll, however it does not occur to portfolio based on Jll. In the share based on Jll, reversal AAR portofolio of the winner share is higher them reversal AAR portofolio of the loser share. In the share based on non J/1, reversal AAR portofolio of the winner share is lower than reversal AAR porrfolio of the loser share. Result of paired sample t test to hypothesis of difference of abnormal return of the winner-loser share in the formation and test periods shows statistically there is significant difference between average AAR during formation and test periods. This strengthens the anomaly of reversal average AAR portofolio during formation and test periods. Result of independent sample t test to toward hypothesis of abnormal return difference between the winner and the loser share during test period shows that there is no significant difference between abnormal return in portfolio based on 111, meanwhile there is significant difference to portfolio of share based on non Jll. For portfolio of share based on Jll, this strengthens result of previous test, where overreaction occurred to the winner and loser share. Result of independent sample t test toward hypothesis of difference of abnormal return share based on J/1 and non J/1 in the test period shows that there is no significant difference to average abnormal return portofolio of the loser share based on ill and non Jll. Different condition occurred to portfolio of the winner share where in the holding period 1 and 3 months occurred significant difference to average abnormal return respectively to portfolio 6 and 8 share and 4 and 6 share. For holding period 3 months, significant difference of average abnormal return occurred to portfolio of the loser share. Conclusion of this research is that assumption portfolio of the share based on J/l has higher resistance against overreaction resistance is not evident, especially for the loser share. Result of statistic test in hypothesis 3 shows that there is significant difference over average AAR during formation period and test for all portfolio based on JII and non Jfl. Result of statistic test to hypothesis 5 shows more specifically that similarity to the existence of overreaction of portfolio based on Jll and non JII occurred to the loser share, marked with non existence of significant difference statistically over average AAR of the loser share based on J/1 and non ill. Further, this condition also indicates that the shares based on non Jl/ is more accustomed to bad news information and being sensitive to good news information, particularly in case the shares based on non JII being the loser share. In addition, it is obtained that the more shares in portfolio, the more non reversal conditions occurred. This will strengthen importance of portfolio diversification strategy for the more stable investment result. Especially for portfolio of the winner share, the longer holding period in portfolio, the average difference of abnormal return between formation period and test period would be bigger with possibility biggest abnormal return would be obtained over investment to the shares with holding period of 3 months. For portfolio of the loser share there is no trend like the winner share. Key words: winner-loser anomaly, abnormal retun, portfolio, Jakarta Islamic Index, 111, holding period, good news, bad news, reversal, overreaction.

Kata Kunci : Anomali winner-loser,Abnormal return,Portofolio,Jakarta Islamic Index,Holding period,Good news,Bad news,Pembalikan,Reaksi berlebihan