[ABSTRAK
Tesis ini membahas kemampuan penggunaan candlestick chart untuk memprediksi trend harga saham dan memperoleh return positif. Candlestick chart adalah salah satu metode analisis teknikal yang telah lama digunakan dalam kontrak futures beras di Jepang sejak tahun 1700-an. Sampel data yang digunakan dalam penelitian ini adalah data harian empat komponen harga saham (open, high, low, close) emiten LQ45 yang terdaftar di Bursa Efek Indonesia (BEI) selama periode Januari 1995 sampai Desember 2013. Data yang digunakan dalam penelitian ini diperoleh dari website thompsonreuters. Metode statistik yang digunakan untuk menguji hipotesis, meliputi one mean t-test dan binomial test dengan SPSS 16. Hasil penelitian menunjukkan bahwa rata-rata return dari tiap pola candlestick tidak sama dengan nol, dan nilainya akan semakin baik jika memperhatikan trend dan volume saham diperdagangkan. Probabilitas perolehan return positif tertinggi mencapai 76% untuk volume transaksi harian diatas 4.000.000 lembar. Untuk memperkuat hasil penelitian, dilakukan pengujian atas data sub periode krisis dan non krisis serta sub sampel kelompok harga saham dan ukuran perusahaan.

---

ABSTRACT
This thesis explored predictive power of candlestick chart in the Indonesian Stock Market. Candlestick chart is one of technical analysis that has been used in Japanese rice market in the 1700s. The data used in this thesis is four daily price (open, high, low, close) of 45 stock which listed in Indonesian Stock Exchange within January 1995 untill December 2013. The data used is acquired from thompsonreuters website. The statistical method used to test the hypothesis are one mean t-test dan binomial test which is run by SPSS 16. The result indicate that average return from each candlestick pattern is not zero, and the result is more significant when considering trend and minimal daily volume of stock traded. Probabillity to acquired highest positive return is 76% for data with minimal daily volume 4.000.000 shares. As an attempt to further analysis, we also test data for sub period which is crisis and non crisis and also sub sample test based on range of stock price and company size.;This thesis explored predictive power of candlestick chart in the Indonesian Stock Market. Candlestick chart is one of technical analysis that has been used in Japanese rice market in the 1700s. The data used in this thesis is four daily price (open, high, low, close) of 45 stock which listed in Indonesian Stock Exchange within January 1995 untill December 2013. The data used is acquired from thompsonreuters website. The statistical method used to test the hypothesis are one mean t-test dan binomial test which is run by SPSS 16. The result indicate that average return from each candlestick pattern is not zero, and the result is more significant when considering trend and minimal daily volume of stock traded. Probabillity to acquired highest positive return is 76% for data with minimal daily volume 4.000.000 shares. As an attempt to further analysis, we also test data for sub period which is crisis and non crisis and also sub sample test based on range of stock price and company size., This thesis explored predictive power of candlestick chart in the Indonesian Stock Market. Candlestick chart is one of technical analysis that has been used in Japanese rice market in the 1700s. The data used in this thesis is four daily price (open, high, low, close) of 45 stock which listed in Indonesian Stock Exchange within January 1995 untill December 2013. The data used is acquired from thompsonreuters website. The statistical method used to test the hypothesis are one mean t-test dan binomial test which is run by SPSS 16. The result indicate that average return from each candlestick pattern is not zero, and the result is more significant when considering trend and minimal daily volume of stock traded. Probabillity to acquired highest positive return is 76% for data with minimal daily volume 4.000.000 shares. As an attempt to further analysis, we also test data for sub period which is crisis and non crisis and also sub sample test based on range of stock price and company size.]