Pairs trading consists of two stages: pairs selection and trading based on the selected stock pairs. The first stage is the key to higher returns. Among existing pairs selection methods, pairs trading based on Engle-Granger cointegration has been proven to be superior. However, the cointegration approach is computationally expensive and brings high latency which may greatly affect the returns. We drastically simplify the Engle-Granger cointegration algorithm, and propose a hardware accelerator for the modified algorithm. In the experiment of selecting stocks of length 5000, our hardware design is more than 1290× faster than CPU and 190× faster than GPU.