I think QE was necessary, and I think the Fed has done a good job implementing it.  But one thing we have to be honest about is that it WILL, if it is done right, lead to inflation past the target at some point in the future.

There are many ways we can view the process of the Fed expanding its balance sheet and buying up a mix of Treasury and risky private debt:

1. They are increasing the money stock directly, which if it gets spent will increase the money supply stimulating the economy,
2. They are buying up long term debt, pushing down longer term interest rates directly,
3. They are taking on risk at a point in time where people may be too “risk averse”.

All these considerations have their pro’s and con’s – but none of them directly imply that the Bank will deviate from its inflation target in the future.  Instead it is a fourth, and arguably the most important, part of QE policies that ensures they will do so – their use of the Fed balance sheet to influence their own future behaviour, making the “optimal path” for monetary policy time consistent.

The situation

Think of it this way, the Fed currently sets policy to minimise deviations in output from its natural rate, and inflation from its target level.  They weight the different outcomes, and set policy consistent with their dual mandate.

However, the Fed doesn’t want to make losses on its balance sheet either.  By expanding its balance sheet, and making it that the returns on the balance sheet are negatively correlated with interest rates (all other things equal) they will enter another factor into their “loss function” which will cause them to keep interest rates lower then they would have under normal “discretion”.

This would normally be a bit weird, but times are weird thanks to a “liquidity trap”.

They are currently facing a zero lower bound on interest rates, they can’t change the current interest rate to achieve their targets.  However, future interest rates also impact upon the current state of the economy, so if they can lower future interest rates they can stimulate the economy now.  The constraint they have is that, in the future, they won’t take into account the beneficial impact of interest rate settings on the past – so the market will recognise this and the Fed’s ability to get the US economy out of a rut will be constrained.  This implies that the “optimal path” from this point in time is time inconsistent – if they say they will do it, they will have the incentive to deviate from this path in the future.

Here, the Fed can use the balance sheet mechanism to ensure that the “future Fed” has to keep interest rates lower.  It is a precommitment mechanism!  So they load up the balance sheet, and in the future they keep rates lower in order to limit their losses.  The market knows this is time consistent, and so prices in lower future interest rates.  Hazzah, we are our of our liquidity trap!

And future inflation

Now, given that future interest rates are lower than a purely “dual mandate” Fed would have targeted we know for a fact that inflation will also be higher – at least until a normal balance sheet is restored.

To me, this is incredibly similar to price level targeting (or even in some form of NGDP targeting) albeit an indirect version.  The Fed is implementing a mechanism that makes them take into account previous “deviations” from their target – when in the past they would be solely forward looking.

I wonder if the Fed actually wanted to introduce price level targeting and/or NGDP targeting, but instead had to use the QE mechanism to sell it politically.  Or, did the Fed think it lacked the credibility to target the price level and had to physically constrain itself – much like I have to with alcohol

Note:  For anyone with AER access, “Credible Commitment to Optimal Escape from a Liquidity Trap: The Role of the Balance Sheet of an Independent Central Bank” by Jeanne and Svensson in Mar 07 gives the best run down of this I’ve seen.

Update:  As I’m pro the policy there is a point I should make clear.  Copy and pasting from email correspondence I said:

The optimal path involves violating their dual mandate in the future on the upside, just by less then they would violate it now on the downside.

That is important – it isn’t that they are suddenly ignoring their mandate.  The ZLB just puts them in a tough place, and from the view of a relatively static loss minimisation perspective (or a simple forward looking trade-off perspective) it looks like a violation of the mandate as they are taking the past into consideration.  However, if we look at it in conjunction with the past the actions appear more sensible.

Update 2: Mankiw has discussed similar issues – but without the call to time inconsistency.