Problems on the model with trend

Estimation ML/Bayesian estimation
8
  1. No, with dy=y-y(-1)+ctrend; you have in steady state dy=ctrend;. Thus, dy has a non-zero mean (a mean estimated from the data), which is fine.
  2. Again, it depends on the trends, the production function and preferences. With unit elasticites there is often not an issue. For more details on preferences that are unproblematic, see Modeling trend growth in DSGE, what utility function should be adopted? - #2 by jpfeifer
  3. No, the mean value in the data is equal to the trend growth rate plus a term depending on the convergence to the BGP. It will overstate the trend growth.
  4. The impact of what? What exactly is the research question?