This paper provides an empirical time series analysis of the daily returns of ten Kuwaiti stock data. We have analyzed the data using simple linear regression model and Box-Jenkins auto regressive integrated moving average (ARIMA) model. We have also investigated the possibility of seasonality over five days of the week (Sunday to Thursday when the transaction takes place). No seasonal pattern is observed. Box-Jenkins approach is also not found suitable for forecasting these indices. However, in some cases, we observe that auto correlation function (ACF) and partial auto correlation function (PACF) are significant at lag one which contradicts the random walk hypothesis. Overall simple regression model is found to give the best forecast.