On QT − remarks by Huw Pill

Remarks

Thanks to the Pictet Research Institute for their invitation to participate in their inaugural symposium.

It is a great honour and pleasure to engage with such a distinguished group of academics and practitioners on contemporary issues of monetary policy.

Those issues are many, inter-related and complex.

While I don’t expect to attract much sympathy in making this remark, we live in challenging times for monetary policy makers like me!

This morning, I will focus on one such monetary policy issue: the pace at which central banks should shrink their balance sheets via ‘quantitative tightening’ (QT).

QT entails reducing the stock of bonds (typically government bonds) held by central banks for monetary policy purposes whose purchase was financed by the creation of reserves. It represents the unwinding of previous quantitative easing (QE).

I hope you will permit some parochialism here, as I will discuss this issue in a UK context.

I appreciate that this discussion may seem a little ‘niche’ or narrow to an international audience. But it sits within a broader web of institutional and policy questions relating to monetary policy and central bank independence. These are topics of global concern.

Last week the Bank of England’s Monetary Policy Committee (MPC) announced a slowing in the pace of QT from £100bn per year to £70bn per year. This decision was taken on the grounds that a number of factors – larger term premia on long-term government bonds, greater global economic policy uncertainty, and weaker demand for longer-term government debt stemming from structural changes in the UK bond market – may have increased the risk that QT would disrupt on market functioning.

There are certainly risks in this direction and I recognise that these have to be addressed pragmatically by the UK authorities.

Nonetheless, I dissented from the majority MPC view on this issue, favouring instead a continuation of the £100bn annual pace of QT that has been implement in recent years. This dissent reflects the high weight I place on maintaining continuity and consistency in the MPC’s approach to QT.

In the remainder of my remarks this morning, I will expand on my reasons for this dissent.

In August 2021, the MPC set out a set of principles to govern the implementation of QT in its Monetary Policy Report^{footnote [1]}. Paraphrasing that document (in the form adopted in recent MPC communication) ^{footnote [2]}:

• The Committee has a preference to use Bank Rate as its active policy tool when adjusting the stance of monetary policy;
• Sales would be conducted so as not to disrupt the functioning of financial markets, and only in appropriate conditions; and
• To help achieve that, sales would be conducted in a gradual and predictable manner over a period of time.

These principles have served the MPC well.

They have helped keep QT “in the background” (to quote an oft-used phrase).

This has helped to support the clarity and simplicity of our monetary policy communication around the “active” Bank Rate instrument. The approach has also been honest and transparent: the MPC has neither used QT as an active monetary policy instrument nor discussed doing so.

Clarity, simplicity, honesty and transparency are cornerstones of effective central bank communication. Our experience with QT has demonstrated this, representing a (perhaps rare) example of where they work to reinforce on another rather than exhibiting trade-offs among themselves.

To be clear, operating “in the background” does not mean (at least to me) that QT has no effect on yields (and wider financial conditions). The evidence suggests there has been a modest impact.^{footnote [3]}

What permits QT to operate “in the background” is the scope for Bank Rate (as the “active instrument”) to establish a policy stance that delivers inflation sustainably at target given the impact of QT on yields. With Bank Rate away from its effective lower bound and able to change in either direction, this is the environment in which QT has operated in recent years.

Operating within our established principles has allowed the market to price the impact of QT and has thereby allowed the MPC to set Bank Rate to achieve the inflation target given the impact of QT – as well as a multitude of other factors – on the yield curve, bank behaviour and wider financial and credit decisions.

Just to emphasise, taking Bank Rate decisions given market pricing of yields and bank credit and loan choices is an entirely standard part of the policy setting process.

Operating gradually and predictably – and making QT announcements early in the autumn policy process – allows QT decisions to be digested and priced by the market before the Bank of England staff develop the analyses and forecasts upon which its November Monetary Policy Report and concurrent Bank Rate decision are based.

I recognise that the world does not stand still. Principles established in 2021 (before I joined the MPC) are not sacrosanct. And decisions taken on the basis of those principles may need to evolve as the wider environment evolves.

• Market conditions have changed, with greater pressure on long rates as demand from defined benefit (DB) pension schemes recedes.
• Fiscal conditions have changed, with the UK’s debt management office (DMO) issuance rising as challenges to public finances intensify.
• Risks to market functioning have increased – not just in the interbank money market (where QT’s impact on the outstanding stock of central bank reserves may be most relevant) but also in the gilt repo markets central to non-bank financial institution (NBFI)’s activity that is gaining in importance relative to bank finance in the transmission of monetary policy.

These are all legitimate concerns that deserve thoughtful analysis and responses.

But I am not convinced that these changes are either being driven by QT or that QT should address their implications.

On that basis, I favoured a different balance between slowing QT to manage risks to market functioning versus maintaining consistency and continuity in the implementation of QT (and thus running down the portfolio of government debt held for monetary policy purposes more quickly) relative to the majority of the MPC.

At best, slowing QT is a temporary and indirect palliative, likely dominated in terms of effectiveness by other tools. And the danger exists that measures to treat the symptoms simply allow the underlying drivers to continue for longer and in greater force, making the eventual denouement more painful all around.

For me, both structural challenges to public debt management and concerns about core market functioning are better dealt with via other means than QT.

There are both institutional and effectiveness aspects to this assertion.

It is when responsibilities are blurred that policy maker accountability and independence are most at risk.

And I have faith in the tools that have been designed and introduced to support market functioning (such as the financial stability operations initiated in October 2022 in the face of the LDI / mini budget episode or the contingent NBFI gilt repo facility^{footnote [4]}). If QT (which was not designed with that function in mind) is portrayed as the first line of defence in this regard, I worry about the risk of undermining them.

I hope I have convinced why I voted to maintain the pace of QT at £100 billion over the coming year. It was driven by a motivation to provide continuity and consistency in the MPC’s approach, particularly as gilt market developments had been predominantly unrelated to QT.

And with that, I am happy to take your questions.

The text has benefitted from helpful comments from Saba Alam, Andrew Bailey, Jamie Bell, Michael McLeay, Arif Merali, Ben Nelson, Adrian Paul, Amar Radia, Andrea Rosen, Vicky Saporta, Kavya Saxena, Martin Seneca, and Tim Willems for which I am most grateful.