Pairs trading merupakan salah satu cara yang paling umum digunakan dalam penerapan statistical arbitrage. Di Indonesia currency pairs trading sudah banyak dilakukan, namun pairs trading dengan menggunakan saham belum terdengar luas. Strategi yang dapat dipakai bermacam - macam, semua berorientasi untuk mencari hubungan antar variabel yang paling kuat. Data harga saham yang digunakan adalah saham yang masuk dalam IHSG dan sudah IPO sebelum periode pembentukan pairs. Data diperoleh dari Bloomberg Terminal. Pairs dibentuk dengan strategi distance method, cointegration, dan hurst exponent. Dengan dua periode, pembentukan pairs menggunakan data masing masing 5 tahun, kemudian diberlakukan backtesting selama 1 tahun. Hasil menunjukkan bahwa metode cointegration dan hurst exponent menghasilkan untung pada kedua periode, sedangkan pairs yang dibentuk dengan strategi distance method mengalami kerugian pada kedua periode. Namun setelah dilakukan simulasi tahunan dan dilakukan uji signifikansi, hurst exponent dan cointegration memberikan hasil yang sama baiknya. / Pairs trading is one of the most widely used strategy in statistical arbitrage implementation, Currency pairs trading is pretty common in Indonesia, but pairs trading using stocks is rarely heard of. There are various strategies that can be put into use in performing pairs trading, yet they all of share one common purpose - that is to look for one strongest relationship between variables. The data that were used in this thesis are stock prices that are already IPO before the pairing period that had been included in The IHSG. Data was collected from The Bloomberg Terminal. The pairs were selected using distance method, cointegration, and hurst exponent strategy. By implementing two periods, pairs choosing processed 5 years historical data from each stock. Afterward, a backtesting was conducted for one year The result showed that the implementation of cointegration and hurst exponent method gave positive return on both period, while pairs that were choosen using distance method strategy had given negative return on both periods. However after applying yearly simulation, hurst exponent eventually gave better returns.