We perform stock market analysis using data from multiple stocks (including index funds and companies). Our approach is based on both statistical modeling and LSTM neural networks. For statistical modeling we use autocorrelation plots to examine trends in data and root mean squared error (RMSE) as our key performance indicator. Using the LSTM neural network we design a regression model for forecasting and a classifier to predict whether to buy, hold or sell stocks at any given day. Finally, we explore the LSTM regression model’s ability to generalize to multiple stocks, as well as its usage for multi-day forecasting.