IEO evaluation of the Bank of England’s approach to quantitative easing

Foreword from the Chair of Court

Maintaining price stability is at the heart of what the Bank of England does. For the first decade of its existence, the Bank’s Monetary Policy Committee (MPC) was able to achieve that stability through changes in one policy tool – Bank Rate. But in responding to the global financial crisis, Bank Rate was cut to close to zero and the MPC, like policymakers around the world, had to turn to other – less familiar – tools to meet its aims. In 2009, the MPC began a programme of quantitative easing (QE), through the Bank’s Asset Purchase Facility, intended as a temporary measure to support the economy in the aftermath of the crisis. But a decade on from its introduction in the UK, QE has become bigger, broader and more persistent than expected. That is why in 2019 the Bank’s Court commissioned this evaluation of the Bank’s asset purchase programme. And, although not a key focus of this report, the further expansion of QE in 2020 as the MPC responded at pace and scale to the Covid-19 crisis, has only made this review more timely.

The Independent Evaluation Office (IEO) was asked to carry out an end-to-end evaluation of the QE rounds between 2009 and 2016. That covers the Bank’s understanding of how QE works, tool design and implementation, governance and risk management and communication. Its findings are described in this report.

The report shows that the Bank has delivered asset purchase programmes effectively. There has been impressive and collaborative staff input to deliver innovative programmes at pace over a number of rounds of QE – in particular in the midst of the global financial crisis in 2009 and again as the Bank responded to the 2016 vote to leave the EU. The Bank’s researchers made a valuable contribution to the growing literature on the effects of QE – especially in the early stages. The Bank has developed strong governance and risk management to underpin delivery of QE. The bespoke technical infrastructure that facilitates purchases has proved resilient, overseen by experienced staff. And the Bank has honed its communications on what is a complex tool. More generally, there is good evidence of the Bank learning and adapting as the programme has expanded. The recommendations in this report are intended to build on and encourage such learning.

Those recommendations fall under three themes. The first considers how the Bank can continue to advance and apply its technical understanding of QE, given that important knowledge gaps remain. The second looks to ensure that the governance and implementation of QE remain fit for the future, recognising the complexity and risk considerations of asset purchases. And the third focuses on building public understanding and trust in QE, a tool which remains poorly understood and often contentious. Taken together these three themes and accompanying recommendations can help to reinforce QE’s role in the Bank’s toolkit. And the lessons from the first decade of using QE can also inform the Bank’s approach to future policy design.

At our 11 December meeting, Court welcomed the Bank’s commitment in taking forward these recommendations. We will monitor their implementation as part of the IEO’s follow-up framework.

Brad Fried, Chair of Court
January 2021

Executive summary

In July 2019, the Court of Directors (the Bank’s board) commissioned its Independent Evaluation Office (IEO) to conduct an evaluation of the Bank’s approach to its Asset Purchase Programme, commonly known as quantitative easing (QE).

QE was first launched in the UK in March 2009 in the midst of the financial crisis, when the Monetary Policy Committee (MPC) announced £75 billion of asset purchases, financed by the issuance of central bank reserves. With Bank Rate having been cut to close to its effective lower bound, the aim of QE was to provide further monetary stimulus, in order to boost economic activity and meet the inflation target in the medium term.

Since then, QE has become significantly larger in scale, broader in type and more persistent than was initially expected. As of November 2020, the MPC has announced QE purchases totalling £895 billion, equivalent to over 40% of annual UK GDP. The vast majority of these purchases have been of UK government debt. The QE portfolio also includes around £20 billion of corporate bonds. QE has proved more persistent than initially expected, with repeated additional purchase rounds and no reductions in the stock of purchases.

This growing role for QE reflects both significant shocks and long-term, structural trends. The impact of the financial crisis proved longer lasting than initially anticipated. It also became clear that global ‘equilibrium’ interest rates – those consistent with full employment and stable inflation at target over the medium term – had fallen. That meant that average policy rates needed to be lower than in the past in order to meet the MPC’s aims. In turn, this increased the chance of Bank Rate being near its lower bound, such that other tools would be needed in the face of negative shocks. Moreover, the UK economy has faced significant shocks – the euro-area crisis, the vote to leave the EU and Covid-19 – since then.

The expanded role for QE underlies the importance of this evaluation. QE should no longer be seen as a transient, ‘unconventional’ crisis response. Instead, it is now an established part of the monetary toolkit that has been used in the UK and in many other countries in response to a range of shocks. It is likely to continue to play a key role for central banks for years to come, at least while equilibrium interest rates remain low. Nevertheless, the use of QE has sometimes been challenged. For example, there has been significant public debate about the possible spillovers associated with QE, with a particular focus on the potential for distributional effects.

This evaluation was commissioned in 2019, with the aim of covering QE from its inception to 2016 – the last round of purchases at the time. While the evaluation was in train, the MPC voted to expand purchases further in response to the severe downturn associated with the Covid pandemic. The IEO does not comment on live policy decisions. But we have drawn on the experiences of the most recent round of purchases, particularly where there has been clear learning from previous episodes or where issues that we had previously identified were brought into sharper relief.

The IEO was tasked with a broad evaluation of the Bank’s approach to the end-to-end QE process. Our focus is not on reviewing estimates of QE impact – there is already a large literature on that. Instead, we have assessed the Bank’s understanding of the QE transmission mechanism and its potential spillovers, the Bank’s approach to designing and operationalising QE, the associated governance and risk management framework and the Bank’s QE communications. Overarching all this, we have assessed the Bank’s learning over the past decade about a new tool. An important focus of our recommendations is about how this learning can continue to be fostered and embedded.

Across our evaluation criteria, we found a wide range of positive evidence on the Bank’s evolving support for the MPC’s QE decisions. The Bank has significantly advanced its understanding of QE over the past decade and has made a good contribution to the international QE evidence base. Since 2009, the Bank has also developed its QE design choices along a number of dimensions in response to new shocks. A common finding across successive rounds of QE is that the Bank has excelled at delivering at pace and under pressure, drawing on very effective staff input from multiple areas. The governance frameworks associated with QE have been strengthened and have proved flexible. Market participants and journalists told us that the Bank’s QE communications have been reasonably good overall, given the complexity of the tool.

The nature of QE as a new tool, with a rapidly expanding role, has naturally also brought a range of new challenges, many of them common to other central banks conducting QE. It was a significant shift to move the focus of monetary policy from a short-term interest rate (Bank Rate) to a quantity of purchases (QE). Establishing and communicating the economic rationale for QE – which is to stimulate the economy by lowering longer-term interest rates – has not been straightforward. The nature of the QE transmission mechanism remains uncertain and the Bank’s emphasis has evolved over time. It has been difficult to communicate to the public about a tool that operates indirectly via financial markets and which was launched during a financial crisis. The institutional structure underpinning QE’s implementation is also complex, reflecting the risks associated with large-scale asset purchases. As the size and persistence of QE has grown, so has the importance of learning about how it works, ensuring its robust implementation and building public understanding of the tool. These challenges motivate our recommendations. Table 1 summarises our recommendations, which we arrange into three broad themes.

Since 2009, the Bank has significantly advanced its understanding of QE. Nevertheless, a decade on, there remain open debates about how QE works in different states of the world, its broader interlinkages and its potential limitations. These debates are not surprising and are not unique to the Bank – they reflect the global state of knowledge about a relatively new tool. They can, however, matter a lot for designing the most effective QE programme and should therefore be prioritised. There is considerable value in sustained investment in learning about QE, particularly given its more persistent role in the monetary toolkit. At times over the past decade, and accepting the considerable demand on its resources, the Bank arguably underprioritised such investment work and lacked a structured plan for delivering it. Our recommendations therefore focus on prioritising the outstanding knowledge gaps, ensuring new evidence continues to closely inform QE design choices and fostering the internal and external QE debate. These recommendations are mutually reinforcing.

There are multiple dimensions to designing a QE programme, requiring the effective interaction of many stakeholders. Our review of the associated governance, risk management and infrastructure found that they have been tested at times, but overall have proved fit for purpose. Our recommendations consider how to safeguard QE’s successful delivery in the future. That includes ensuring that QE governance arrangements are clear and well understood. We suggest regularly raising awareness of the cash transfer arrangements between the Bank’s Asset Purchase Facility (APF) and HM Treasury (HMT). In light of the full indemnity HMT provides to the APF, any accounting profit or loss is transferred to HMT each quarter. To date, that has meant a series of positive flows to HMT, totalling around £110 billion. But given that the APF may need to receive cash transfers as QE is unwound, it is important that this scenario continues to be widely understood by relevant stakeholders, ahead of time. We also recommend that the Bank continues to regularly review its internal understanding of where QE design responsibilities lie and the principles behind the relative roles of the MPC and the Executive. That is particularly important given QE’s expanded and probably persistent role as a core part of the monetary policy toolkit. Finally, in light of QE’s expanded role, the Bank may wish to consider further investment in the operational and risk management infrastructure that underpins QE delivery, alongside its priorities for other capital projects.

Good communication to the public supports both the accountability and the effectiveness of the Bank of England. With QE now a core part of the monetary policy toolkit, the public’s trust in and understanding of the tool is important for the Bank’s mission. We recognise the communication challenges that QE has faced, given it is a complex tool with perceived side effects. Feedback on the Bank’s overall communications have been positive and the Bank has seriously engaged with analysis of potential spillovers, albeit relatively slowly. Nevertheless, QE remains a poorly understood monetary policy tool for much of the public and for some its use has been contentious. In particular, there remain strong views about QE’s potential distributional side effects. The absence of climate considerations in the Bank’s approach to its corporate bond purchases has also been challenged, with some arguing that a more positive environmental signal could be sent. Our recommendations are designed to help the Bank better engage with these issues. To continue to build trust and understanding, we suggest that the Bank develops more accessible, layered communications on QE. Recognising the nature of public debate about QE, we also recommend embedding a structured approach to engage with the potential spillovers of any new tool. In order to deliver these, we think the Bank may wish to consider embedding a more strategic approach to QE communications.

Our report focuses on QE. Nevertheless, the Bank may wish to consider whether our findings could be applied to other elements of the policy toolkit.

The evaluation was conducted by a dedicated project team reporting directly to the Chair of Court.^{footnote [1]} The IEO team benefited from feedback and challenge from a Bankwide senior-level advisory group (including Bank Governors and an external member of the Bank’s MPC). Guy Debelle (Deputy Governor, Reserve Bank of Australia) and Ricardo Reis (A. W. Phillips Professor of Economics, London School of Economics) provided support and independent challenge to the team and reviewed and endorsed the findings in this report.

The report was approved for publication by the Chair of Court at the December 2020 Court meeting.

1: Context for the evaluation

In the United Kingdom, as in most advanced economies, monetary policy makers have typically used their ability to change short-term interest rates to help manage fluctuations in demand in the economy and hence inflation. When the financial crisis hit in 2008, the Bank of England cut Bank Rate sharply. By early 2009, it was close to its effective lower bound and with little scope to reduce short-term interest rates further, the Bank then had to consider alternative, so-called ‘unconventional’, tools to support the economy.

In January 2009, the Chancellor announced that the Bank would set up an Asset Purchase Facility (APF), indemnified by HM Treasury (HMT). Initially, the APF was used to purchase high-quality private sector assets, in order to increase the availability of corporate credit.^{footnote [2]} A provision was also made that allowed the MPC to use the APF for monetary policy purposes.

In March 2009, the Monetary Policy Committee (MPC) announced that it would use the APF to buy £75 billion of assets, with the aim of increasing nominal spending growth to a rate consistent with meeting the inflation target in the medium term. It noted that as well as funding any private sector asset purchases under the APF, the Bank would also buy UK government debt (gilts) in the secondary market and that it was likely that the majority of the overall purchases would be of gilts. This type of policy – commonly known as Quantitative Easing (QE) – had been introduced in Japan in 2001 and had subsequently been adopted by the US Federal Reserve as part of its crisis response. Since 2009, a number of central banks have used asset purchases as part of their monetary policy toolkit. Annex 1 provides a summary of international programmes.

QE was introduced in the UK to provide temporary support to the economy. The impact of the financial crisis was, however, longer lasting than initially anticipated. It also became clear that global ‘equilibrium’ interest rates – those consistent with full employment and stable inflation at target over the medium term – had fallen.^{footnote [3]} That meant that average policy rates needed to be lower than in the past in order to meet the MPC’s aims, and increased the chance of Bank Rate being near its lower bound, such that other tools would be needed in the face of negative shocks. Moreover, the UK economy has faced significant shocks – the euro-area crisis, the vote to leave the EU and Covid-19 (Covid) – since then. In response, UK QE has been expanded over four further phases (Figure 1.1).

In the initial phases of UK QE, the MPC provided stimulus through large-scale purchases of longer-term UK government debt. Since then, the vast majority of UK QE purchases have continued to be of gilts. But the MPC has also purchased some private sector assets for monetary policy purposes: the Corporate Bond Purchase Scheme (CBPS), introduced in 2016, saw the MPC purchasing bonds issued by companies that make a material contribution to the UK economy as part of the QE programme.

Figure 1.1: Bank of England QE programmes and selected policy interventions since 2009 ^(a)

Footnotes

• (a) Monetary policy and credit easing policies since 2009. For more information about Bank Rate decisions and Forward Guidance, please see respective MPC minutes. Further information about other schemes can be found here: APF Purchases of Commercial and Corporate Bonds, Funding for Lending Scheme, Term Funding Scheme and Term Funding Scheme (SMEs).

Relative to initial expectations, UK QE has therefore become:

• Much bigger in size: by November 2020, the Bank had announced purchases totalling £895 billion.
• Broader in scope: in 2016, the programme expanded to include Corporate Bond purchases, with £10 billion bought in 2016–17, and a further £10 billion in 2020.
• More persistent: in 2009, the Bank noted that asset sales could be undertaken as part of a tightening in monetary policy as the economy recovered. But that point has not yet been reached.

Overall, QE has transitioned from being a transient, unconventional crisis response, to a persistent part of the monetary policy toolkit. With structural factors putting continued downward pressure on equilibrium interest rates globally, Bank Rate is more likely to come up against its effective lower bound in the event of economic shocks. QE is therefore likely to continue to play a key role in the monetary policy toolkit for years to come.

The remainder of this section summarises the Bank’s approach to QE and our approach to this evaluation.

1.1: The Bank’s approach to QE

Objective and transmission of QE

The objective of QE is to provide monetary stimulus, helping the MPC to meet its inflation target. If the MPC projects inflation to fall below its 2% target, it can use QE to boost demand in the economy, and hence inflation. At a high level, QE involves buying assets from the private sector – financed by the creation of central bank reserves. That is intended to lower longer-term interest rates and, in turn, encourage spending on goods and services, boost economic activity and employment and put upward pressure on prices. As set out in Box A, there are a number of potential channels through which QE supports demand and inflation. Purchases of corporate bonds have a similar impact to gilt purchases, but may also reduce liquidity premia in sterling corporate bond markets, and stimulate issuance and hence corporate borrowing.

Institutional design and governance of the Asset Purchase Facility

The Bank conducts its asset purchases through a separate fund – the APF. The APF is funded by a loan from the Bank at Bank Rate (financed by the creation of reserves). Importantly, the APF is fully indemnified by HMT from any losses arising out of or in connection to the facility. As a separate entity, the APF has its own Directors (who are the Bank’s Executive Directors for Markets, Banking, Monetary Analysis and Finance) and publishes annual accounts, audited by the National Audit Office. Although indemnified, the Bank risk manages the APF on behalf of HMT. The UK approach to delivering asset purchases through a separate indemnified fund is different to that taken by other major central banks, which – reflecting their individual governance and funding arrangements – have typically used their own balance sheets to conduct purchases.

HMT is a key stakeholder in the governance of QE. Initially, HMT indemnified the Bank for up to £150 billion of purchases. Each time the MPC has decided to expand the size or composition of the facility, the Governor has requested an expansion in the value of the assets the APF is authorised to purchase from the Chancellor. The institutional design ensures that the MPC remain operationally independent and fully accountable.

Design and implementation of QE by the Bank

Unlike Bank Rate, there are several potential dimensions to asset purchase decisions – including selecting the quantity, maturity and type of asset – and the purchase of some assets may carry higher risk than others. The governance and risk oversight of QE is therefore a key element of policy design.

The relative responsibilities for the MPC and the Bank’s Executive are outlined within a concordat, designed to cover the broader Sterling Monetary Framework. At a high level, the MPC is responsible for determining the amount of monetary stimulus needed. The Executive are responsible for operational decisions on how that stimulus is delivered and for risk management.

The MPC votes on a stock of gilts and corporate bonds within the APF. So long as the MPC has voted to maintain or increase the stock of purchased assets, the Bank ensures that the proceeds of any maturing assets are reinvested in the APF. Those reinvestments are communicated via the MPC minutes and market notices. Given the scale of gilt holdings in the APF, gilt reinvestments are carried out regularly. For the much smaller stock of corporate bonds, the Bank reinvests periodically when the required reinvestment has reached a sufficient size to conduct an effective auction programme.

For corporate bonds, the Bank purchases sterling investment grade bonds, issued by private non-financial companies that make a material contribution to the UK economy. The assessment of the materiality of the contribution to the economy made by individual businesses is conducted by the Bank’s risk management area.^{footnote [4]} Operationally, purchases are designed to deliver a balanced portfolio of bonds across eligible issuers and sectors, so as not to influence the allocation of credit to particular companies or sectors of the economy.

The Bank’s QE operations are all in the secondary market – it does not buy gilts or corporate bonds directly from their issuers. The Bank offers to purchase gilts and corporate bonds in reverse auctions, which have been designed to take into account the features of each market. These are carried out through its bespoke Btender platform, with specific sets of firms able to participate in its different auctions. Further details of the Bank’s operations are available in the Bank of England Market Operations Guide.

Given its activities in the gilt market, the Bank has established a set of operational arrangements with the UK Debt Management Office (DMO) – which has responsibility for debt and cash management for the UK government and hence the issuance of gilts in the primary market. These establish a clear delineation of roles and ensure smooth functioning of the gilt market. For example, the Bank does not offer to purchase gilts issued by the DMO within one week of their issue. In August 2009, the Bank and DMO also introduced a gilt lending facility to relieve any undesirable frictions in the functioning of the market in specific gilts arising from the Bank’s purchases. And the Bank has made clear that should asset sales become appropriate it will liaise with the DMO in order to minimise interference with their issuance programme.

Internal organisation of QE delivery

The successful delivery of QE relies on input from staff around the Bank.

• Research and analysis on issues such as the effectiveness of QE and modelling its transmission is carried out by Monetary Analysis, Markets and the Research Hub. Colleagues in Financial Stability and the PRA input on the linkages with financial sector balance sheets and the International Directorate bring a global perspective. Together with analysis on the economic outlook, this analysis feeds into the MPC’s policy decision.
• Responsibility for operationalising a decision to purchase assets lies with the Markets area, who consider the approach to purchases and the parameters to ensure smooth market functioning.
• Decisions on new tools and operational choices are scrutinised by the Bank’s risk management functions, which advise the Bank’s Executive on any implications for financial and non-financial risk (such as reputational risks). Legal experts are also consulted on relevant issues.
• At the point of implementation, the Bank’s dealers carry out the auctions and liaise with counterparties; trades are processed by the middle and back office areas in Markets.
• QE decisions are communicated through the MPC’s minutes and the Monetary Policy Report (MPR),^{footnote [5]} with colleagues in the Communications Directorate helping to deliver those messages outside the Bank. The Bank’s Agents also play a key role in communicating the Bank’s policy tools in their regions. 

1.2: Approach to our evaluation

Our evaluation is broad in scope, covering the Bank’s approach to the end-to-end QE process. We based it around four evaluation criteria:

1. Does the Bank have a clear view on the QE transmission mechanism and the overall impact of QE, grounded in the available evidence?
2. Does the Bank have an effective process for designing and operationalising QE?
3. How appropriate is the governance and risk management framework for QE?
4. Does the Bank effectively communicate QE?

Our focus is not on reviewing estimates of QE impact. There is already a large literature on that, with key findings summarised in Box B. Instead, our focus is on the Bank’s overall understanding of QE, the application of that to the implementation of QE in practice and the Bank’s capacity to learn given QE’s expanded and persistent role.

This evaluation was commissioned in 2019, with the aim of covering QE from its inception to 2016 – the last round of purchases at the time. While the evaluation was in train, the MPC voted to expand purchases further in response to the severe downturn associated with the Covid pandemic. The IEO does not comment on live policy decisions. But we have drawn on the experiences of the most recent round of purchases, particularly where there has been clear learning from previous episodes or where issues that we had previously identified were brought into sharper relief.

The evaluation has drawn on a wealth of evidence. This includes around 150 interviews with current and ex-policymakers and Bank staff, academics, think tanks and journalists. We have also reviewed internal and external research, analysis and documentation. And we have carried out textual analysis of both Bank publications and media coverage of QE. Annex 2 provides more details on our remit, scope and method.

2: Continuing to advance and apply technical understanding of QE

The Bank has significantly advanced its understanding of QE over the past decade. In 2009, when QE was first launched in the UK, there was limited international evidence to draw on. Inevitably, that meant there was considerable uncertainty about the effects that QE might have. Since then, a large evidence base has developed (see Box B).

We find that the Bank’s research and analysis has made a notable contribution to the QE knowledge frontier. Chart 2.1 shows the number of QE-related research publications – and the associated number of citations – from the Bank of England over time and compares that to some other central banks. Given the relative size of the institutions, we think the Bank has had a good ‘footprint’ on QE research, particularly with early papers around 2011/12. Our discussions with academics and economic practitioners outside the Bank corroborated this positive assessment (see Box C).

Chart 2.1: Bank of England contribution to QE literature ^(a)

Number of papers and RePec citations for selected central bank empirical QE literature

Panel 1: Number of QE-related publications

Panel 2: Number of RePec citations

Footnotes

• Sources: RePec CitEc project and Bank calculations.
• (a) This chart is based on a selected sample of 84 QE-related empirical papers published by central banks. Panel 1 shows the number of papers published each year between 2009 and 2019. Panel 2 shows the total number of RePec citations of those papers (by year of publication) as at 2 December 2020. We base our sample of papers on a survey of central bank literature from the Bank of England, European Central Bank and the US Federal Reserve. We select papers published with at least one author who worked at these institutions at the time of publication. Our sample focuses only on empirical QE research (rather than papers covering theoretical perspectives). We cross-check our sample against references in the following QE meta-studies: Borio and Zabai (2016), Bhattarai and Neely (2016), Haldane et al (2016), CGFS (2019) and Fabo et al (2020). Our cut-off for publication is December 2019. Citations usually build over time so recent papers will tend to have fewer citations than older ones. Given the large body of QE literature and several alternative sample criteria, this list has been compiled on a best efforts basis and should be treated as indicative rather than exhaustive.

We find that the Bank has excelled at delivering new rounds of QE under pressure in response to varied shocks. This has relied on very effective staff input, which feedback consistently suggests has been collaborative, nimble and impressive. Staff from around the Bank have worked together to design and operationalise QE at pace.

Nevertheless, we have heard that the Bank could benefit from continuing to develop its thinking on the use of QE. This remains important even when not actively considering QE expansion in the near term, in order to improve preparedness for future shocks. At times over the past decade, accepting the considerable demand on its resources, the Bank arguably underprioritised such investment work and lacked a structured plan for delivering it.

This matters, because a decade on there remain open debates about how QE works in different states of the world, its broader interlinkages and its potential limitations. These debates are not surprising and are not unique to the Bank – they reflect the global state of knowledge on a relatively new tool.^{footnote [12]} They do, however, have implications for how best to design an effective QE programme.

There are several dimensions to QE decisions, including selecting the quantity, type and maturity of assets to buy, determining the purchase pace and the approach to reinvestment. The Bank has innovated on these QE design features, in response to new challenges. As design choices are flexed, it is important that the Bank continues to clearly link its evolving understanding of how QE works to how it does it. Externally, this may also improve understanding of the MPC’s reaction function.

Our assessment is that there would be considerable value for the Bank in systematic and sustained QE development work. The theme of this section is therefore continuing to advance and apply technical understanding of QE. Our recommendations are summarised in Figure 2.1. We suggest prioritising the outstanding knowledge gaps and ensuring that evolving understanding continues to closely inform practical QE design choices. Internally, we recommend a regular forum to discuss QE’s role in the event of a big shock. We also suggest that the Bank updates its technical audience on its latest thinking on QE and proactively fosters external engagement with that. Our recommendations are mutually reinforcing.

Figure 2.1: Summary of recommendations

We are mindful that investment work of this kind is subject to resource constraints and competing priorities. Moreover, we recognise that our recommendations might involve some rebalancing of resource away from near-term issues towards longer-term strategic policy questions. Nevertheless, we think there is considerable value in sustained investment in learning more about QE, particularly given its more persistent role in the monetary toolkit.

These recommendations focus on QE. The Bank may wish to consider the extent to which they could also be usefully applied to other elements of the policy toolkit.

2.1: A prioritised QE work plan

Our first recommendation is to produce a prioritised work plan for future research and analysis to fill QE knowledge gaps. It is for the Bank to decide on the key elements of that plan in light of resource constraints and existing commitments. But to aid that, this sub-section notes topics that were raised in our discussions with the Bank’s experts and external stakeholders (see Box C for further detail on externals’ views).

A common finding was that continued analysis and modelling of the QE transmission mechanism would be valuable. As discussed in Box A, QE is thought to operate through several potential channels. The Bank’s emphasis on different channels has evolved over time and a decade on there is not a settled view on which channels matter the most.

Open questions about the QE transmission mechanism are not surprising and are not unique to the Bank. QE is a complex tool and there are a relatively small (albeit growing) number of observations globally, from which to learn. Moreover, the role of different transmission channels is likely to depend on the economic context – so-called ‘state-contingency’ (for further detail, see Bailey et al (2020)). This may have important implications for effective QE design. For example, the relative weight placed on signalling versus portfolio balance channels might affect which assets it is most effective to buy and how best to communicate QE to maximise its impact. Similarly, the importance attached to the liquidity channel may affect judgements on when QE can be most effective and whether the pace of purchases matters. Considerations of this kind might also inform the degree of emphasis placed on the quantum of QE purchases versus the intended impact on long-term interest rates. Drawing out the implications of these different elements of the transmission mechanism for effective QE design, interactions with other tools and noting the key uncertainties would be beneficial.

There was also appetite from Bank experts and external stakeholders for further analysis on the ultimate impact of QE on GDP and inflation. As set out in Box B, the macroeconomic effects of QE are naturally harder to pin down than the immediate impact on financial conditions. Nevertheless, there may be opportunities to glean further evidence on these effects, as more QE observations accrue.

Feedback also supported further analysis of the interlinkages between QE and financial stability. This could build on the Bank’s existing work on the potential implications of QE for financial institutions’ balance sheets, which we found effectively draws on expertise from around the institution. The interactions between QE and financial stability may vary depending on the time horizon considered. For example, in the medium term, potential risks associated with ‘search for yield’ behaviour may be considered. In the near-term, the potential benefits of QE for market functioning were illustrated by recent Covid-related interventions.

Further consideration of potential monetary-fiscal interlinkages was noted as another topic that the Bank may wish to include in a QE work plan.^{footnote [13]} That is consistent with the much larger and more persistent effect of QE on central bank balance sheets than initially expected.

Finally, longer-term issues raised included exploring the potential limits of QE, including how that might interact with very low long-term interest rates, and analysing the implications of QE unwind, well in advance of any decision to reduce the stock of QE. Alongside the Bank continuing to advance its own understanding of QE, it was noted that continuing to assess how external expectations of QE were set was also important.

When considering its QE work plan, we would encourage close interaction between the Bank’s policy and research teams. This is consistent with the Bank’s Agenda for Research (BEAR), the first topic of which is the monetary policy toolkit. It is clearly important that the Bank maintains and develops the human capital required for advancing understanding of the policy toolkit.^{footnote [14]}

2.2: The rationale and evidence supporting practical QE design choices

Our second recommendation is that as the Bank continues to learn about QE, it maintains a summary of the rationale and evidence supporting each of its practical QE design decisions, to clearly link its understanding of how QE works to how it does it.

There are several elements to QE design (see Figure 2.2 for an illustration). Clearly, a key feature of any QE intervention is the quantity of assets purchased. Decisions also need to be made over: the type of asset class to purchase – for example, whether government or private sector debt; the maturity of the assets; and the pace of purchases. Another design choice is whether and how to reinvest cash flows as assets mature.

Figure 2.2: Illustration of elements of QE design decisions

Since 2009, the Bank has developed its QE design choices along each of these dimensions. For example, while the vast majority of QE purchases have been of gilts, corporate sector asset purchases were also included with the launch of the CBPS in 2016 and its extension in 2020. The Bank has also flexed the pace of its asset purchases. For example, given the nature of the Covid shock in March 2020, the initial pace of gilt purchases in response was particularly rapid, peaking at more than double that seen in previous rounds of QE in the UK.

Given the range of potential QE design choices, there may be benefit in maintaining a framework that considers the conditions under which different elements might be flexed. This would help guide the design of any QE response to future shocks. It might also provide a natural opportunity to circle back to previous design decisions, testing the underlying rationale as new evidence emerges.

One case study relates to the CBPS, where feedback suggests a lack of clarity about when and why the Bank buys corporate bonds. A more systematic framework linking the intended CBPS transmission channels to its role in response to different types of shock might be useful. Such an assessment might also help to anchor CBPS design choices. For example, we believe that a cost benefit analysis on the merits of maintaining the CBPS stock through reinvestment may be warranted.

2.3: Regular forum to discuss QE’s role in the event of a big shock

Our third recommendation is that the Bank provide an internal forum to allow regular MPC discussions on how QE (as part of the wider monetary policy toolkit) would be used in the event of a big shock. The aim would be to facilitate impactful QE debate ahead of time, drawing on continued learning and also steering future work plan priorities.

Feedback suggests that the Bank at times over the past decade under-prioritised strategic monetary policy projects. In part, that reflected resource constraints, given many pressing policy challenges. Nevertheless, further work on the policy implications of sustained low equilibrium interest rates (so-called ‘lower-for-longer’) may have been warranted sooner. This might have included additional assessment of QE’s role. Instead, we found that the Bank’s work on QE tapered off somewhat from around 2013. With hindsight, continued analytical investment on the impact and transmission of QE during this period could have brought forward some of the work required for the relaunch of QE in 2016.

In contrast, in recent years the Bank has conducted very effective policy contingency planning in response to known events such as the possibility of a no-deal Brexit. Feedback suggests that the Bank’s preparations for the potential risks associated with Brexit also proved useful when considering the policy response to Covid.

The Bank may wish to consider mechanisms to ensure that similar contingency discussions continue in the longer term. We believe that such discussions remain important, even when there is no large shock on the horizon and when an extension of QE may seem to be a distant prospect.

We found some support for a regular forum to consider ‘big picture’ monetary policy questions. As an illustration, this might ask the MPC and relevant teams about the adequacy of the monetary policy toolkit to respond to a large output gap.^{footnote [15]} We think that such a forum would provide a mechanism to discuss important longer-term issues related to QE. These might include: QE’s role in providing significant future stimulus; the associated risk management trade-offs; state contingencies and potential limitations of QE; and the interaction of QE with other tools.

Such a forum might naturally foster a culture of challenge and longer-term analytical investment. This could reinforce our previous recommendations by encouraging the Bank to test the latest evidence on QE, consider the implications for policy design and, in turn, identify future work streams.

2.4: Update technical audience and foster external engagement

Our fourth recommendation is that, where appropriate, the Bank considers updating its technical external audience on its latest collective thinking on QE and proactively fosters external engagement with that.

The Bank has put out a raft of research, analysis and commentary on QE over the past decade. As an illustration, Chart 2.2 provides a count of QE-related publications by the Bank of England that we have identified, arranged by publication type. Since 2009, we estimate that the Bank has published around 40 speeches, 32 papers, 12 articles, 9 Bank Underground posts and 7 Report boxes materially related to QE.

Chart 2.2: Count of QE-related publications by Bank of England, by publication type ^(a)

Footnotes

• (a) ‘QE-related publication’ defined by a combination of thresholds for the frequency of QE terms and judgement. Data cut-off is 5 November 2020.
• (b) Includes Staff Working Papers, Discussion Papers and External MPC Discussion Papers.

There is, however, a lack of published material on the Bank’s collective view on QE, particularly recently. For example, we heard that there was not much agreed material to draw on when writing about QE in official publications such as the MPC minutes. In part, this may reflect the fact that there is understandably not a settled collective view on exactly how QE works. Moreover, we have heard that the Bank has become increasingly open in welcoming a diversity of views and the associated debate.

We think that a collective update from the Bank on its latest thinking on QE might serve a number of purposes. First, feedback from some external stakeholders suggests that the Bank might usefully clarify some elements of its QE reaction function (see Box C). Second, we note that a technical update from the Bank on QE could be a useful source for other policy communications, such as the MPC minutes. Third, an external update on QE might also provide a helpful focal point around which to organise internal debate, consolidating our first three recommendations. Finally, this could foster continued external engagement on the topic. Where there is not a firm collective view, this would be an opportunity to be open about key outstanding uncertainties and their implications. Setting out the priorities for the QE work plan might also help to re-catalyse external debate, including with academics and peer central banks. As an example, we heard that the Bank’s 2011 QE conference was very successful in bringing together early views on QE.^{footnote [16]}

The appropriate timeline for any external update on QE is for the Bank to determine, in light of current challenges and associated resource constraints. But the Bank may wish to consider bringing together the latest evidence from the recent ‘QE5’ round of purchases, building on Bailey et al (2020). Further ahead, we believe the MPC should continue to share – at a moderate frequency – its best collective view on how one of its key policy tools supports the economy. We note, for example, that the FPC publishes a policy statement for each of its powers of direction to explain how it plans to use them and why.

3: Ensuring that the governance and implementation of QE remain fit for the future

The Bank’s approach to the governance of QE has evolved steadily over the past decade. This partly reflects the demands of a larger and more complex set of asset purchases than foreseen when QE was launched in 2009. It also reflects the evolution of the Bank as an organisation and changes to the governance of its overall balance sheet and operations.^{footnote [17]}

In this section, we focus on three key elements that underpin QE governance and implementation:

• external governance arrangements with HMT;
• internal governance to facilitate QE design decisions and associated risk assessment; and
• the delivery of QE through market operations and associated infrastructure.

Overall, we find that the external governance arrangements for QE are well designed and have functioned effectively. As set out in Section 1.1, the Bank’s APF was set up as a wholly-owned subsidiary of the Bank, which is indemnified by HMT. The institutional structure has allowed appropriate oversight and accountability and has provided the Bank with the necessary operational independence to deliver QE. Effective working relationships between the Bank, HMT and the DMO have been built over time. The Bank should continue to maintain and embed these working relationships, through continued regular engagement on both current and forward-looking topics of mutual interest.

Internally, the Bank has strengthened its framework for QE governance. As discussed in Section 2.2, there are several dimensions to QE decisions with implications for QE’s impact, the appropriate operational approach and risk. Implementing QE therefore relies on the interaction of a range of stakeholders. In response, the Bank has developed clear terms of engagement and respective decision-making remits between the Bank’s Court, Executive and MPC, which are set out in an MPC Concordat.^{footnote [18]} The risk management framework for the APF has also become increasingly comprehensive over time, aided by an internal review in 2015.^{footnote [19]}

The technological infrastructure underpinning the Bank’s operations and risk management has also developed over time. This infrastructure, run by experienced staff, has been instrumental in effectively monitoring and delivering larger and more complex QE operations over the past decade.

Given QE’s role as a core part of the Bank’s monetary policy toolkit, our recommendations consider how to safeguard QE’s delivery in the future. This includes ensuring that QE governance arrangements are clear and well understood. We suggest regularly raising awareness of the cash transfer arrangements between the APF and HMT. We also recommend that the Bank continues to regularly review its internal understanding on where QE design responsibilities lie and the principles behind that. Finally, the Bank may wish to consider further investment in the operational and risk management infrastructure that underpins QE. These recommendations are summarised in Figure 3.1.

Figure 3.1: Summary of recommendations

3.1: Raising greater awareness of the implications of APF cash transfer arrangements

Our first recommendation is that the Bank regularly raises awareness of APF cash transfer arrangements, and their implications during unwind, to key senior stakeholders and externally.

The APF is a subsidiary of the Bank, and is indemnified by HMT. This means that any financial losses as a result of asset purchases are borne by HMT, and any gains are owed to HMT. The institutional set up has important implications for the Bank and public sector balance sheets.

The assets in the APF generate income and their value will fluctuate over time. When the Bank engages in asset purchases for monetary policy purposes, it increases the loan to the APF financed by newly created central bank reserves. In turn, the APF purchases the required assets (currently gilts and corporate bonds) from the private sector. These assets generate income for the APF through, for example, coupon payments. This income is used to fund the interest payments on the APF’s loan from the Bank (set at Bank Rate) and any administrative costs. These flows need not be equal, so over a given period it is possible for the APF to make an accounting profit or loss. Moreover, if the proceeds on maturity of an APF asset are different to the original cost (which will tend to be the case if it was not bought at exactly par value), then this will also result in an accounting profit or loss to the APF.^{footnote [20]} Panel 1 in Figure 3.2 illustrates these cash-flow arrangements.

In 2012, a change was made to cash management arrangements between the APF and HMT. An exchange of letters set out that there would be a quarterly transfer of net cash flows accruing to the APF. Any net profit made by the APF (adjusted for forthcoming known expenses) at the end of each quarter is transferred to HMT, with net losses resulting in a transfer from HMT to the APF.^{footnote [21]}

Since QE was launched in 2009, interest rates have fallen and stayed low, and the MPC has not reduced the stock of QE assets. The Bank has re-invested the proceeds from maturing assets to maintain the QE stock. The income generated from APF asset holdings has been sufficient to cover interest payments on the loan from the Bank, administrative costs and any shortfalls arising when proceeds from maturing assets (bought above their par value) have been below their price at the time of purchase. There have therefore been a series of positive net flows from the APF to HMT, cumulating to around £110 billion as of October 2020 (Panel 2 in Figure 3.2).

Figure 3.2: APF cash-flow arrangements

Footnotes

• Sources (Panel 2): Office for National Statistics and Bank calculations.

Although net flows to HMT have been positive so far, the APF can also experience net cash shortfalls, which would require transfers from HMT to the APF. The exchange of letters makes clear that HMT will make these return payments on a timely basis, in order to limit any risks to the Bank of England balance sheet. The Bank’s analysis, published in 2013 (McLaren and Smith), notes that while such shortfalls are likely to be large under many scenarios, their potential size and timing are uncertain and will vary depending on factors such as the total size and composition of asset holdings, the future path of Bank Rate and the path of any asset sales.

Given the one-way nature of these flows between the APF and HMT so far, however, there could be reputational risks if any large future cash-flow reversals took any stakeholders by surprise. We heard that future scenarios for APF cash flows are regularly discussed at working level between the Bank and HMT, as part of the risk management process. And the issue is discussed periodically with senior stakeholders.

Given the potential for fading institutional and public memories, it is important that arrangements to manage a reversal of cash flows remain well understood by relevant stakeholders, ahead of time. We therefore recommend that the Bank continues to ensure that key stakeholders are aware of the implications of these cash transfer arrangements. As part of this, the Bank may wish to share regular updates of APF cash-flow simulations with senior stakeholders. External awareness could also be raised, for example, with references in relevant Bank speeches and through an update of the analysis in McLaren and Smith (2013). Until September 2017, the Bank published a spreadsheet illustrating future cash-flow scenarios on its website. Updating and re-publishing these scenarios, alongside explanatory notes, may help interested parties better understand these arrangements.

3.2: Reviewing internal understanding of the principles of the MPC Concordat

Our second recommendation is that the Bank reviews internal understanding of what the principles of the MPC Concordat mean for QE design responsibilities in practice.

The Bank’s internal oversight and decision-making framework for QE has strengthened through time. As discussed above, implementing QE relies on the interaction of multiple areas, spanning policy, operational and risk management functions. In response, the Bank has developed a framework of engagement for key senior stakeholders, which include the Bank’s Court, Executive and MPC. This is outlined in an MPC Concordat, which was last updated in June 2018, following an internal review. Box D summarises how the Bank organises its QE governance and risk management.

Stakeholders noted that the overall principles underlying these governance arrangements are well shaped and have facilitated decision-making. At a high level, the concordat notes that decisions relating to the selection of policy tools (such as asset purchases) with the primary intention to affect monetary conditions and the amount of stimulus needed to achieve the inflation target are within the MPC’s purview. The Executive’s role is to select the operational framework to design and implement MPC decisions, and to assess the risks to the Bank’s balance sheet, as delegated by Court. Feedback suggests that the high-level nature of these principles is helpful and enables flexibility in times of crisis, when quick decisions are needed under uncertainty.

We heard of some occasions over the period since 2009 where it was not necessarily clear whether some dimensions of QE design were principally for the Bank’s Executive or for the MPC. For example, the Bank’s decision in 2012 to change its QE maturity buckets for gilt purchases led to a degree of internal debate at the time. These issues have been navigated successfully. We also heard that there are advantages in having some flexibility, rather than attempting to codify the full range of potential QE design and implementation responsibilities. Given QE’s expanding role and complexity, the Bank should remain alert to the possibility of future ‘grey areas’ in internal governance responsibilities.

Overall, we believe that more could be done to raise internal understanding of what the principles of the MPC Concordat mean for evolving QE design responsibilities in practice. While there have been some staff efforts to do this so far, we believe the Bank could do more to review and systematise such discussions in the future. Given the range of potential QE design choices (as discussed in Section 2.2), having clarity around the framework for decision-making can continue to facilitate effective governance in future. Moreover, a common message from stakeholders was that well-functioning internal governance is likely to remain particularly important while Bank Rate remains close to zero, and the MPC has to use historically ‘unconventional’ policy tools. That is because such tools are likely to have additional implications for financial and non-financial risk on the Bank’s balance sheet.

Some stakeholders also noted that there may be merit in considering a light-touch update of the Concordat. For example, to reference APF indemnity arrangements as per the Financial Relationship principles set out in the 2018 Bank-HM Treasury Memorandum of Understanding.

3.3: Prioritising further investment in operational and risk management infrastructure

Our third recommendation is that the Bank considers prioritising further investment to make improvements to operational and risk management infrastructure.

The Bank has developed internal systems and technological infrastructure to implement and oversee its asset purchase programmes. This includes an in-house electronic tendering system (Btender) and systems and processes to help staff monitor the risk implications of QE. Bank staff have effectively responded to the increasing challenges of an expanding QE portfolio, requiring more operations and more complex risk oversight and reporting. That said, stakeholders have said that further investment in the associated technical infrastructure could improve and support the Bank’s operations over the long term, and may also reduce dependence on key experienced staff.

Btender has been used for the Bank’s QE operations since 2009. We heard that the system was well designed and has been operationally resilient and reliable through the past decade. However, the technological infrastructure is now facilitating a much larger and more persistent QE portfolio than initially envisaged in 2009. Btender requires significant input to run and maintain. External counterparties face occasional idiosyncratic issues when interacting with it. So far, staff efforts have been effective in mitigating any negative consequences, which have at times needed ad-hoc changes to the infrastructure. We also note that the Bank has been considering a project to explore investment in more permanent infrastructure improvements (such as greater automation) to Btender, but this project continues to be on hold given competing priorities and budget constraints.

Similarly, the Bank’s financial risk monitoring and management infrastructure relies on number of complex systems and processes, which require significant staff input to run and extract relevant information. For example, the process to calculate the sensitivity of APF risks to various future cash-flow scenarios continues to require significant resource. As noted in Section 3.1, a regular and reliable update of this type of scenario analysis is likely to remain an important aspect of future APF risk management. The Bank has designed a project called ‘Financial Risk Architecture Simplification’ that is intended to build more integrated and automated systems. These systems would also support the implementation of a recommendation in the 2015 Risk Review. This project has also been on hold, given competing priorities and budget constraints.

In light of this, we recommend that the Bank considers prioritising further investment to improve the operational flexibility and resilience of its platforms, alongside its priorities for other capital projects, to help ensure effective delivery of future rounds of asset purchases and other market operations.

4: Building public understanding and trust in QE

Good communication with the public supports both the accountability and the effectiveness of the Bank of England. With QE now a core part of the monetary policy toolkit, the public’s trust in and understanding of the tool is important for the Bank’s mission.

Journalists and other external stakeholders that we talked to were positive about the Bank’s overall communications with the public. The Bank now talks to the public much more than when QE was launched in 2009.^{footnote [22]} We heard the Bank’s approach to communication compares well to that of other central banks.

External stakeholders also recognise that communicating about QE is challenging. It is a complex tool. QE relies on indirect transmission channels via financial markets and has a less direct impact on retail financial products such as mortgages than Bank Rate (see Box A). The Bank also first deployed QE around the same time that the government recapitalised failed banks during the financial crisis and then tightened fiscal policy. QE may therefore be associated in peoples’ minds with their views on these other policies. Finally, repeated shocks to the UK economy have prevented the Bank unwinding QE; it has only ever expanded it. As a result, any potential winners and losers from QE may appear persistent.

As a result, QE remains a contentious tool in both the UK and abroad. For example, in a 2018 survey, fewer than 40% of UK MPs thought using QE again would be advisable. This reflects lively debates about QE’s potential side effects:

• The potential distributional effects of QE are the key area of contention. Bank work (see Section 4.2) suggests that the overall effect of monetary policy and QE on standard measures of income and wealth inequality has been small. But by reducing long-term interest rates, QE – alongside the more general low rate environment – is seen by many to hurt savers. Meanwhile to the extent that QE raises the prices of assets, ownership of which is fairly concentrated, many worry that it increases inequality. Consistent with this, text analysis shows that the share of UK press articles about QE that discuss it in the immediate context of distributional words (eg ‘savers’, ‘the rich’, ‘inequality’) has increased markedly since 2011 (Chart 4.1). The share hit around a third of QE articles after a speech by Theresa May in 2016 criticised the distributional impacts of QE.
• Another area of contention is the absence of climate considerations in the Bank’s approach to buying corporate bonds.^{footnote [23]} The Bank’s Agents and our press analysis suggest that the ‘Green QE’ debate is less broad-based than the distributional debate. But the topic still elicits strong views from some of the public.
• The public is also unclear about the extent to which QE is, or should be, used to finance Government borrowing. Given the UK’s post-Covid fiscal position, a lack of public clarity on monetary financing could undermine the Bank of England’s independence in the future.^{footnote [24]}

Chart 4.1: Co-location of distributional and QE words in UK press articles about QE ^(a)

Footnotes

• Sources: Dow Jones Factiva and Bank calculations.
• (a) Articles are from City AM, Daily Mail, Daily Telegraph, Evening Standard, Guardian, i, Independent, Mail on Sunday, Metro, Mirror, Observer, Sunday Times and The Times (where City AM, i, Sunday Times and The Times only contain articles from 2010 onwards). Data are only available to end-2019 so do not include QE5.
• (b) Percentage of QE articles that contain a distributional term in +/-20 words of a QE term. ‘QE articles’ contain at least one term from both a list of QE terms and a list of Bank of England terms. These term lists were compiled by the Independent Evaluation Office.

Our recommendations are designed to help the Bank rise to the challenge of communicating about QE and to better engage with these debates. Figure 4.1 summarises our three recommendations. Boosting public trust in and understanding of a complex tool like QE is of course a challenge. But the emerging literature on central bank communications suggests some lessons for doing so (see Box E). The Bank can aim to better support its sophisticated audiences in conveying the Bank’s QE messages to the broader public. In doing so, it can help the person on the street to be at least as familiar with the positive objectives of QE as with its potential spillovers.

Figure 4.1: Summary of recommendations

4.1: Develop more accessible layered communications on QE

Our first recommendation is that the Bank develops more accessible layered communications on QE. The Bank ‘layers’ its key publications, with each layer addressing a different target audience. As Box E sets out, there is evidence that this can build comprehension, trust and engagement with central bank messages. Accessible material explaining a complex tool such as QE is particularly valuable because of the scepticism and even distrust that has grown around its effects.

Since 2009, the Bank has developed its technical explanation of QE through Inflation Report boxes, Quarterly Bulletins and speeches by senior Bank officials. A KnowledgeBank page distils these into an accessible explanation for the public.

There is, however, further scope to improve the Bank’s explanation of QE. External feedback and the literature on central bank communications suggest a more effective explanation could be:

• Simple and story based: Rather than explaining QE in terms of ‘injecting’ or ‘printing’ money, QE could be framed as a continuation of conventional monetary policy that pushes down long-term interest rates to support the economy.^{footnote [25]} This is a more intuitive description of the QE transmission mechanism set out in Box A and provides a simple story of how QE works, using concepts more familiar to the public.^{footnote [26]} Framing QE as a change in an interest rate rather than the creation of a quantity of money may also reduce the perception that QE is a transfer of wealth to the rich or to banks.
• Relatable: It could include a greater focus on the macroeconomic impacts on economic activity, inflation and especially jobs. Ideally, this would be grounded in real world examples. In early communications, the Bank’s desire to show that QE was having an impact led to a focus on asset prices – a key initial part of the transmission mechanism. This fuelled the perception that QE chiefly helps those holding financial assets.

The Bank could also provide accessible responses to the broader debates around QE. On the basis of questions asked at Citizens’ Panel events and recent online Governor Q&As, the KnowledgeBank page can only answer about a fifth of questions on the public’s mind (Chart 4.2). Echoing this, some public-facing Bank staff felt they lacked accessible materials setting out the Bank’s position on issues such as Green QE.

While the Bank has ad hoc internal ‘lines-to-take’ on these questions, it could consider building them into a published FAQ to increase the chance that internet searches for ‘QE & inequality’ or ‘Green QE’ would be directed to a Bank web page.^{footnote [27]} And rather than just setting out the counterfactual that welfare would be lower without QE, we also heard that the Bank could be more explicit about why alternative policy designs are not possible.

Chart 4.2: Audience questions about QE at Citizens’ Panels and Governor Q&As ^(a)

Footnotes

• (a) Questions categorised by the Independent Evaluation Office. These questions were asked at Citizens' Panel events held around the UK between November 2018 and February 2020 or submitted to online Governor Q&As held in July and October 2020.
• (b) This includes designs that would bypass the financial system such as People's QE or helicopter money, as well as designs that would target QE’s impact at certain parts of the real economy (eg at manufacturing firms).

4.2: Embed a structured approach to engage with the potential spillovers of any new tool

Our second recommendation is that the Bank embeds a more structured and pro-active approach to engaging with the potential spillovers of any new tool. QE’s potential side effects play a big role in the public debate about the tool. But talking about these issues can be challenging for the Bank because many do not fit neatly in its remit. This recommendation is designed to help the Bank engage on these fronts.

The debate about the distributional effects of QE has proved particularly challenging. Since 2011, a number of MPC speeches have covered this issue and a Quarterly Bulletin article was published in 2012. The Bank’s most comprehensive and impactful work on this topic – a 2018 Staff Working Paper – was published nearly a decade after QE was launched. That work suggests that the overall effect of monetary policy and QE on standard measures of income and wealth inequality has been small, because while QE supports asset prices, it also supports jobs and wages. We heard some praise for the underlying analysis (see Box C). While international evidence on the impact of unconventional monetary policies on inequality is mixed, the paper’s conclusions are in line with the prevailing consensus that looser monetary policy reduces inequality by supporting employment (see IMF IEO (2019) for a recent summary of the literature).

The Bank has struggled to persuade external audiences of its message on distributional effects. Internal and external interviewees described the Bank as ‘defensive,’ while the Bank’s Agents said that some local contacts are sceptical of the analysis. We also heard external criticism about how the Bank presented and interpreted the analysis (see Box C), highlighting a deficit of trust on the issue.

By contrast, the Bank’s communication around Green QE drew less criticism. The absence of climate considerations in the Bank’s approach to CBPS has been loudly criticised by some. But we heard praise for the Bank’s disclosure that the CBPS portfolio was not aligned with Paris climate targets, because it allowed for an open public debate about the potential to ‘green’ QE. The Bank’s Climate Hub has also been open to regular engagement with Green QE proponents. The Bank has now stated that it is considering how to incorporate climate factors into decisions on the mix of financial assets that it holds, while ensuring the policy aims of the relevant portfolios are still met.^{footnote [28]}

Learning from these two contrasting experiences, we think the Bank could benefit from a more structured and pro-active approach to talking about a new policy’s potential side-effects. This could encourage early analysis of side effects, especially distributional ones.^{footnote [29]} Conducting and publishing such analysis shows the Bank takes public concerns seriously, and can underpin a range of communications. Recent strategic reviews by the US Federal Reserve and European Central Bank have featured discussions about increased engagement with distributional and (for the ECB) climate issues.

To maximise impact, the Bank could quickly distil its technical analysis of a tool and its assessment of potential side effects into accessible, layered communications. For example, a simple explanation of the widespread benefits of a tool for the real economy, and especially jobs, is likely to be particularly powerful in countering potential concerns about side effects, before they become embedded.

To build trust in these messages, the Bank may wish to invest further in more open external engagement. For example, we heard that greater staff-level contact – supported by press office – with journalists and external stakeholders around the publication of key analysis could help the Bank to build their understanding. It could also start a constructive two-way dialogue, before entrenched positions start to build distrust.

4.3: Embed a more strategic approach to QE communications

Our third recommendation is that the Bank embed a more strategic approach to QE communications. Building public trust and understanding in a tool like QE probably requires consistent and repeated messages delivered across a range of media. This recommendation is about building the internal processes and tools to achieve that.

Internally, we heard that communications have at times been reactive and something of an afterthought in the policy process. There is appetite for the top of the Bank to set clearer communications priorities to direct and coordinate messaging on QE. An important consideration is how to help individual MPC members – a key voice for the Bank – explain QE in a consistent manner, while voicing their own views on policy. We also heard that the Bank could spend more time building up a constituency of support for QE outside of policy rounds.

Specific processes and tools that the Bank may wish to consider include:

• A forum for the Bank’s senior leadership to discuss communications strategy for QE, including prioritising and co-ordinating messages across digital communications, speeches, direct engagement.
• More analytical capacity to monitor the media to help prioritise and pro-actively engage with public debates (eg Chart 4.3 shows how text analysis can monitor spikes in QE press coverage, while Chart 4.1 is one way of monitoring the content of the coverage).
• Joined up forward planning to allow commissioning of analytical work that can, in good time, support QE communications.

Parts of our recommendation that the Bank adopts a more strategic approach to talking about QE are likely to apply to the Bank’s broader communications.

Chart 4.3: Frequency of QE articles in three major UK newspapers ^{(a) (b)}

Footnotes

• Sources: Dow Jones Factiva and Bank calculations.
• (a) Articles that include at least one term from both a list of QE terms and a list of Bank of England terms. These term lists were compiled by the Independent Evaluation Office. Data are only available to January 2019 so do not include QE5.
• (b) Articles are from The Daily Mail, The Guardian and The Mirror and filtered for the subjects Commodity/Financial Market News, Corporate/Industrial News, and Economic News.

Annex 1: International QE programmes

A range of central banks have used asset purchases as a policy tool. Several have purchased government bonds and high-quality private sector debt as in the UK. Table A1.1 summarises programmes launched internationally since 2001.

The evaluation criteria in Table A2.1 focus on the Bank’s understanding and implementation of QE. Our focus is not on reviewing estimates of QE impact: there is already a large literature on that, with key findings summarised in Box B.

We used several methods to conduct our evaluation as set out in Table A2.2.

The evaluation was primarily conducted between December 2019 and October 2020. This evaluation was initially commissioned to cover QE from its inception to 2016 – the last round of purchases at the time. While the evaluation was in train, the MPC voted to expand purchases further in response to the severe downturn associated with the Covid pandemic. The IEO does not comment on live policy decisions. But we have drawn on the experiences of the most recent round of purchases, particularly where there has been clear learning from previous episodes or where issues that we had previously identified were brought into sharper relief.

The evaluation benefited from input and challenge from IEO-appointed external advisors Guy Debelle (Deputy Governor, Reserve Bank of Australia) and Ricardo Reis (A. W. Phillips Professor of Economics, London School of Economics). They provided support and independent challenge to the team and reviewed and endorsed the findings in this report.

The IEO team was also aided by a Bankwide Senior Advisory Group. This included internal senior representation from Monetary Analysis, Markets, Financial Stability, International, Communications, Legal and Risk directorates. It included the Deputy Governors for Monetary Policy and Markets, Banking and Resolution and an external MPC member.^{footnote [33]} The group provided comment and challenge throughout the project.

Consistent with previous IEO evaluations, the team’s work was run at arm’s length from the business areas and reported directly to the Chair of the Bank’s Court of Directors. As such, recommendations and analysis contained in this report are the sole responsibility of the IEO. The report is based on data and information available as of 5 November 2020, unless otherwise stated.

References

Bailey, A, Bridges, J, Harrison, R, Jones, J and Mankodi, A (2020), ‘The central bank balance sheet as a policy tool: past, present and future’, Paper prepared for Jackson Hole Economic Policy Symposium, 27–28 August 2020. Also now available as Bank of England Staff Working Paper No. 899.

Bank for International Settlements (BIS) (2019), ‘Unconventional monetary policy tools: a cross-country analysis’, Committee on the Global Financial System (CGFS) Papers No. 63.

Bank of England (2018a), ‘Financial relationship between HM Treasury and the Bank of England: memorandum of understanding’.

Bank of England (2018b), ‘Inflation Report’, August.

Bank of England (2018c), ‘The MPC and the Bank’s Sterling Monetary Framework’, Updated in June 2018.

Bank of England (2020a), ‘Annual Report and Accounts 1 March 2019 – 29 February 2020’, pages 43–49.

Bank of England (2020b), ‘APF Annual Report’.

Bank of England (2020c), ‘Bank of England/TNS Inflation Attitudes Survey - May 2020’, June.

Bank of England (2020d), ‘The Bank of England’s climate-related financial disclosure 2020’, June.

Bank of England’s Independent Evaluation Office (2018), ‘Evaluation of the Bank of England’s approach to providing sterling liquidity’.

Bank of England’s Independent Evaluation Office (2019), ‘Evaluation of the Bank of England’s research’.

Bean, C (2009), ‘Quantitative easing – an interim report’, Speech to the London Society of Chartered Accountants, London, 13 October.

Bernanke, B (2020), ‘The new tools of monetary policy: American Economic Association Presidential Address’, The Brookings Institution, 4 January 2020.

Bhattarai, S and Neely, C (2016), ‘An analysis of the literature on international unconventional monetary policy’, Federal Reserve Bank of St. Louis Working Paper 2016-021.

Bholat, D, Broughton, N, Ter Meer, J and Walczak, E (2019), ‘Enhancing central bank communications using simple and relatable information’, Journal of Monetary Economics, Vol. 108(C), pages 1–15.

Binder, C (2017), ‘Fed speak on main street: central bank communication and household expectations’, Journal of Macroeconomics, Vol. 52, pages 238–51.

Borio, C and Zabai, A (2016), ‘Unconventional monetary policies: a re-appraisal’, Bank for International Settlements (BIS) Working Paper No. 570.

Bunn, P, Pugh, A and Yeates, C (2018), ‘The distributional impact of monetary policy easing in the UK between 2008 and 2014’, Bank of England Staff Working Paper No. 720.

Butt, N, Churm, R, McMahon, M, Morotz, A and Schanz, J (2014), ‘QE and the bank lending channel in the United Kingdom’, Bank of England Working Paper No. 511.

Candia, B, Coibion, O and Gorodnichenko, Y (2020), ‘Communication and the beliefs of economic agents’, National Bureau of Economic Research (NBER) Working Paper 27800.

Coibion, O, Gorodnichenko, Y and Weber, M (2019), ‘Monetary policy communications and their effects on household inflation expectations’, University of Chicago, Becker Friedman Institute for Economics Working Paper No. 2019-03.

Coibion, O, Gorodnichenko, Y, Kumar, S and Pedemonte, M (2020), ‘Inflation expectations as a policy tool?’, Journal of International Economics, Vol. 124(C).

Dräger, L, Lamla, M and Pfajfar, D (2016), ‘Are survey expectations theory-consistent? The role of central bank communication and news’, European Economic Review, Vol. 85(C), pages 84–111.

European Central Bank (ECB) (2015), ‘Is the ECB ‘expropriating’ savers?’.

Fabo, B, Jančoková, M, Kempf, E and Pastor, L (2020), ‘Fifty shades of QE: conflicts of interest in economic research’, University of Chicago, Becker Friedman Institute for Economics, Working Paper No. 2020-128.

Federal Reserve (2020), ‘Review of monetary policy strategy, tools and communications’, Board of Governors of the Federal Reserve System.

Gagnon, J (2016), ‘Quantitative easing: an underappreciated success’, Peterson Institute for International Economics (PIIE) Policy Brief No. PB16-4.

Giansante, S, Fatouh, M and Ongena, S (2020), ‘Does quantitative easing boost bank lending to the real economy or cause other bank asset reallocation? The case of the UK’, Bank of England Staff Working Paper No. 883.

Gürkaynak, R (2018), Central Bank Communications paper session discussant, American Economic Association Annual Meeting, 6 January 2018.

Haldane, A (2018), ‘Climbing the public engagement ladder’, Speech given to Royal Society for the encouragement of Arts, Manufactures and Commerce (RSA), 6 March 2018.

Haldane, A and McMahon, M (2018), ‘Central bank communications and the general public’, American Economic Association (AEA) Papers and Proceedings, Vol. 108, pages 578–83.

Haldane, A, Macaulay, A and McMahon, M (2020), ‘The 3 E’s of central bank communication with the public’, Bank of England Staff Working Paper No. 847.

Haldane, A, Roberts-Sklar, M, Wieladek, T and Young, C (2016), ‘QE: the story so far’, Bank of England Staff Working Paper No. 624.

Hall, R and Reis, R (2015), ‘Maintaining central-bank financial stability under new-style central banking’, National Bureau of Economic Research (NBER) Working Paper 21173.

International Monetary Fund (IMF) (2019a), ‘IMF advice on unconventional monetary policies’, Independent Evaluation Office.

International Monetary Fund (IMF) (2019b), ‘The risks and side effects of UMP: an assessment of IMF views and analysis’, Independent Evaluation Office.

Joyce, M (2012), ‘Quantitative easing and other unconventional monetary policies: Bank of England conference summary’, Bank of England Quarterly Bulletin, Bank of England, Vol. 52, No. 1, pages 48–56.

Joyce, M, Lasaosa, A, Stevens, I and Tong, M (2011), ‘The financial market impact of quantitative easing in the United Kingdom’, International Journal of Central Banking, Vol. 7(3), pages 113–61.

Joyce, M, Tong, M and Woods, R (2011), ‘The United Kingdom’s quantitative easing policy: design, operation and impact’, Bank of England Quarterly Bulletin, Bank of England, Vol. 51, No. 3, pages 200–12.

Lagarde, C (2020), ‘The monetary policy strategy review: some preliminary considerations’, Speech at the ECB and Its Watchers XXI conference.

Lamla, J and Vinogradov, D (2019), ‘Central bank announcements: big news for little people?’, Journal of Monetary Economics, Vol. 108, pages 21–38.

Leslie, J, Rahman, F, Smith, J and Gagnon, J (2019), ‘Quantitative (displ)easing? Does QE work and how should it be used next time?’, Resolution Foundation, September 2019.

McLaren, N and Smith, T (2013), ‘The profile of cash transfers between the Asset Purchase Facility and Her Majesty’s Treasury’, Bank of England Quarterly Bulletin, Bank of England, Vol. 53, No. 1, pages 29–37.

Reis, R (2016), ‘Funding quantitative easing to target inflation’, Discussion Papers 1626, Centre for Macroeconomics (CFM).

Reis, R (2019a), ‘Can the central bank alleviate fiscal burdens?’ in: Mayes, D G, Siklos, P L and Sturm, J-E (ed.), The Oxford Handbook of the Economics of Central Banking.

Reis, R (2019b), ‘Central banks going long’, Central Banking, Analysis, and Economic Policies Book Series, in: Aguirre, A, Brunnermeier, M and Saravia, D (ed.), Monetary policy and financial stability: transmission mechanisms and policy implications, edition 1, Vol. 26, Ch. 3, pages 43–81, Central Bank of Chile.

Runge, J and Hudson, N (2020), ‘Public understanding of economics and economic statistics’, Economic Statistics Centre of Excellence (ESCoE) Occasional Paper 03.

Weale, M and Wieladek, T (2016), ‘What are the macroeconomic effects of asset purchases?’, Journal of Monetary Economics, Vol. 79, pages 81–93.

Williams, J (2013), ‘A defense of moderation in monetary policy’, Federal Reserve Bank of San Francisco Working Paper 2013-15.

Winters, B (2012), ‘Review of the Bank of England’s framework for providing liquidity to the banking system’, October.