I have a variable operating income and I want to auto regress it on lagged operating income using the Holtz-Eakin, Newey and Rosen technique. I want to include time dummies to remove effects of heterogeneity. My panel consists of firm level data identified by gvkey for years 2005-2019 identified by fyear. From what i read I should be using xtabond or xtabond 2,but I am not sure how to do it. Could somebody provide any direction? Thanks!