Click-bait of a title, I know. 10.25pm on a Sunday night too. But now that I have scared any potential readers away, I can get down to saying something that I think is quite interesting (without worrying whether it is sufficiently interesting). And it sort of relates to QE.
As a recap, it is worth reflecting (yes, again) on how QE works. While it is only possibly true that no-one knows, it is certainly the case that there is uncertainty and disagreement amongst the principal architects. As Bernanke quipped “the problem with QE is that it works in practice but it doesn’t work in theory”.
I have seen three main competing/ complimentary explanations as to how QE might work which are rather clumsily assembled on the diagram below:
Over the past month the Bank of England has published a couple of empirical studies on what effect a couple of these channels had. The first, on the portfolio balance channel (basically the belief that in buying gilts, the Bank would effectively force institutional investors into riskier assets) which most Central Bankers have relied on in public to articulate the mechanism through which QE translates into easier monetary condition, found that pension funds and insurance companies had increased corporate bond holdings and cash while reducing both gilt and equity holdings versus a counter factual simulation of their behaviour absent QE. This finding doesn’t dismiss the portfolio balance channel, but I read it as a statement that it was pretty disappointing versus the Bank’s expectations.
The second, on the bank lending channel, was pretty dismissive as to any potential empirical support. In the authors’ own words ‘we find no evidence to suggest that QE operated via a traditional bank lending channel’.
This leaves two possible transmission mechanisms: a confidence/ signalling effect, and Everything Else We Don’t Yet Understand (EEWDYU).
One could be forgiven for thinking that Mario Draghi has put in the hours in order to prove that signalling is what it is all about. After all, he delivered the most seismic easing in monetary conditions that Europe has seen in the last decade simply by deploying the phrase “whatever it takes” with a conviction few could doubt. But Europe’s inability to sustain an economic recovery to match its financial market recovery from the depths of the Eurozone crisis puts the idea that signalling is *all that* in doubt.
People seem pretty united around the idea that QE works. So, forgetting how QE might work for a second, let’s recall what QE is. It is basically a massive debt swap, exchanging long-term liabilities (aka gilts) for short-term government liabilities (aka high powered money). This is the case whether we think QE works via the portfolio balance effect, bank lending channel effect, confidence/ signalling effect or EEWDYU. In other words, if QE was worth doing, managing the term structure of the government’s liabilities appears to be a big deal.
Who is responsible for managing the term structure of the government’s liabilities? The UK Debt Management Office – an institution that used to be part of the Bank of England run by charming and capable professionals with whom I might be (I hope only temporarily) unpopular on account of my (very sensible) suggestion that they refinance the War Loan (which I recognise would be an administrative hassle for them although a very good thing for people keen to reduce the national debt and HMT interest payments).
There is also the possibility that QE actually made absolutely no difference whatsoever and has been a massive red herring.
So here’s the thing. If QE works through some mechanism other than a confidence/ signalling effect the rationale for splitting the DMO from the BoE appears defunct. Is it time for a reunion?