Doubts on the Role of Disturbance Variance in New Keynesian Models and Suggested Refinements

Models with rational expectation have become quite popular in macroeconomics, particularly in the context of New Keynesian Models and DSGE models, respectively. These models are useful in many respects; however, they suffer from a serious problem which is discussed here in a very basic version: These models have equations with white-noise disturbance terms with finite variance where the size of this variance is assumed to have no impact on the behavior of economic agents and the equilibrium solution or the steady state values, respectively. This, however, is totally implausible - from a theoretical perspective, a disturbance term with a very large variance, for example, should have a crucial impact on consumption, investment and output, respectively. In the context of the Great Recession and the Transatlantic Banking Crisis as well as in the Euro Crisis - during which one could observe very high volatility of bonds prices pointing to a high variance of disturbance terms - one may therefore raise critical questions with respect to validity of policy recommendations derived from DSGE models. Hence these models should be refined adequately; institutions and regulations have an influence on the variance of white noise disturbance terms and thus various institutional regimes with differences in the variances of these disturbance terms should be discussed. Selected digital expansion and innovation aspects - e.g. related to the Corona shock - also are highlighted.