A few weeks ago I presented a graph that showed the trade-off between a big central bank balance sheet and faster NGDP growth:
Now I'd like to use this framework to discuss why so many people confuse easy and tight money policies. In the following graph, I've contrasted easy and tight monetary policies, over an extended period of time:
In the tight money case, the economy moves from point A to point B. This occurs as the central bank accommodates an increased demand for base money, by doing QE. The demand for base money is rising because the expected future NGDP growth rate is slowing---perhaps due to bad "signaling" from the central bank. This path reflects Japan during the 1990s and 2000s, or the ECB since 2008.
In the other path, an expansionary monetary policy is adopted, but it takes a while before it has any credibility. Initially, the central bank does the same sort of QE as the BOJ did in Japan, causing the Base/NGDP ratio to increase. But in this case there is a subtle difference. Now the quantity of base money is above the money demand curve, resulting in excess cash balances.
Over time, the excess cash balances leads to a hot potato effect, which gradually boosts the NGDP growth rate. As NGDP growth expectations rise, the demand for money (as a share of GDP) begins to fall, and you move from point C to point D.
The point is that the move from A to B looks pretty similar to the move from A to C. You can't just look at variables such as the monetary base (i.e. QE) or nominal interest rates, and ascertain the stance of monetary policy. You need to look at NGDP growth expectations.
How does the central bank insure that you move from A to C, rather than from A to B? There are many options, including currency depreciation, NGDP futures targeting, level targeting, and a "whatever it takes" approach to open market purchases.