I found another excellent work of Michael Woodford on the inclusion of monetary aggregates in modeling inflation dynamics. Again, he shows that money in an engogenous money-supply regime is not needed in order to determine inflation at both high-frequencies and low-frequencies (although there is empirical support for a cointegration of money and inflation at low frequencies). Furthermore he simulates the exact same results as Assenmacher-Wesche and Gerlach (2004) when employing a standard New-Keynesian environment with an endogenous evolving quantity equation relation. What is exciting is the fact that he does not need to reject quantity theory, any kind of monetarist critiques on New-Keyensian models or a denial of a stable money-demand relationship.
Nevertheless, it remains questionable to what extend money can serve as information for the outlook of inflation. According to Woodford, we do not need this second pillar of the ECB because there is only one sole inflation determination, i.e. in line with the economic analysis. Furthermore, Woodford admits at the same time that the information extraction of the analysis of monetary aggregates can be included in an inflation targeting framework.
However, as the excellent Charles Goodhart recently put forward, in a monetary economy with financial frictions (borrowing constraints, credit channel of monetary transmission etc.), there can occur supply shocks to money, i.e. the willigness or the ability of banks to lend to the private sector. This is the reason why an analysis of credit aggregates rather than monetary aggregates may improve the information set of a central bank.
In my view, this is an essential weakness in New-Keynesian models which assume frictionless financial market. The inclusion of risk aversity, income/wealth ratios and further financial frictions would improve New-Keynesian modeling.
