I have an aggregate macroeconomic time series variable, which one would expect to grow exponentially. So, in order to forecast it, I use the stationary time series of its growth rate over time. I am looking at the forecast errors of 2020 Q2, Q3 and Q4, to measure the effect of the covid-19 lockdown on aggregate earnings by industry. In my paper, I want to simplify the table I present by showing the cumulative forecast error over Q2–Q4, instead of each quarter's growth rate forecast error.
Can I add up the standard errors from each quarter, cumulatively?