This paper demonstrates that, despite the current mandate of monetary policy, its final goal (at least for central banks of developed countries) is the control of three main macroeconomic variables — economic growth, employment and inflation, — regardless on actual mandate for this policy. However, the priorities of realization of the final goal may face the imperfection of macroeconomic models and rules of monetary policy, which will make it impossible to control all three macroeconomic variables at the same time. The article proposes a new instrument for monetary policy — aggregate cumulative market imperfection — to optimize macroeconomic variables and stabilize cyclical economic dynamics. The author demonstrates the main competitive advantages of this instrument of monetary policy as compared with typical models of macroeconomic dynamics and simple rules of monetary policy (Simons, Friedman, and Taylor rules). In particular, this instrument is valid for any combination of market conditions, for any economy and for any moment of real time. It can be used simultaneously as: 1) a target of monetary policy; 2) a simple rule of monetary policy correction in the short-run; 3) a reaction function to evaluate a backward connection between the regulator’s actions and the effect of these actions on current economic situation; and 4) an instrument to stabilize cyclical economic dynamics; 5) an instrument to forecast starting (ending) point of recessions and shift in macroeconomic trends. If we can hold the aggregate cumulative market imperfection within a given optimal interval with the help of government regulations (i.e. to target this indicator only) using all possible instruments both of monetary, and (if necessary) of other kinds of regulation policy, we will be able to optimize all three main macroeconomic variables. Optimality of these variables means providing maximum economic growth and employment under comfortable inflation for any combination of market conditions and for any moment of calendar time, which will at the same time stabilize cyclical economic dynamics. In doing so, we will not target each of these three variables separately, that is, it is practically impossible to determine quantitatively their optimal values as they change permanently over time together with the constant change of current combination of market conditions.