The nation’s economy greatly depends on the stock market. A healthy stock market can boost the economy, but if not going well, it can also trigger a recession. It is one of the most volatile markets, it is subjected to sudden changes due to variety of factors including interest rates, supply and demand, inflation, political issues, natural disasters, etc. Therefore, accurate stock market forecasting is crucial in order to help stock customers. Many models which have already been worked on for predicting stock prices are using techniques like ARIMA and LSTM. But hybrid models are now being developed for better outcomes. This paper proposes composition of ARIMA and LSTM for stock prediction and compares the results of ARIMA, LSTM and ARIMA-LSTM hybrid. To test the effectiveness of the hybrid model to others, the model is applied to 5 businesses from various industries. For comparison, different errors are taken into account.