Suppose the Bank of Canada were following a 5% NGDP level-path target. And suppose that actual NGDP was on target, and was expected to remain on target in future. And suppose you were Governor, and one of your advisors gives you some important news. Financial markets are in a bubble, so the prices of financial assets are above their fundamental values. And you know your advisor is always right about this sort of thing (just suppose).
You are an economist. You know there are almost always trade-offs. Keeping NGDP on target is important, but so is financial stability. And there are always diminishing marginal benefits to everything. So it seems rational to sacrifice a little NGDP stability to gain some financial stability at the margin. Who knows, you tell yourself, sacrificing a little NGDP stability in the short run might even help NGDP stability in the long run, because it's harder to keep NGDP on target if financial markets are doing silly things like bubbling and bursting.
But you also believe in transparency, so if you depart from your NGDP target temporarily you must announce that you are doing so.
What would you announce?
Would you announce that NGDP would grow more slowly than 5% temporarily?
Or would you announce that NGDP would grow more quickly than 5% temporarily?
Which of those two announcements would cause interest rates to rise and asset prices to fall?
The Fisher equation in your New Keynesian macro model says that an increase in expected inflation will cause nominal interest rates to rise.
The consumption and investment Euler equations in your New Keynesian macro model say that an increase in expected real GDP growth will cause real interest rates to rise.
Expected NGDP growth is the sum of expected inflation plus expected real GDP growth.
Here is the Governor's recent speech on "Integrating Financial Stability into Monetary Policy".
[Update: just to spell it out, my bottom line for this post is: if we try to 'integrate financial stability into monetary policy', do we really have a clue what we are doing? If not, maybe we shouldn't do it.]