ABERDEEN STANDARD INVESTMENTS – WRITTEN EVIDENCE (QEI0009)
QUANTITATIVE EASING INQUIRY
Executive summary
With Quantitative Easing (QE) first implemented following the 2008 financial crisis, and after the Bank of England’s (BoE) announcement of a further programme, now is an opportune moment to assess its impacts and the legitimacy of continuing to support the UK economy in this way, so ASI welcomes the opportunity to respond to this consultation.
QE is when a central bank purchases long-term government bonds and other assets, which helps to ease financial conditions and provide monetary support even with interest rates close to the effective lower bound. While QE may once have been considered ‘unconventional’ monetary policy, the persistent low interest rates we are expecting mean that QE is likely here to stay and must therefore be seen as part of the Bank’s ‘conventional’ toolkit. There is no reason to think the use of QE up to this point or its continued use in the future will undermine the BoE’s independence.
Evaluating QE’s precise impact over the last decade is complicated by the multiple other economic shocks since the policy’s implementation, however the overwhelming weight of evidence points to it having helped to boost economic activity. Now, after the economic damage and widened inequalities resulting from the pandemic, the current QE programme will be vital to help generate growth, increase wages and reduce unemployment, and we are hopeful that its operation in conjunction with a more supportive fiscal policy than in 2009, will lead to greater benefits.
We explore these points as well as issues around the BoE’s future mandate and QE programmes in more detail below in response to the consultation questions.
Consultation questions
Along with directly setting bank rate, and providing forward guidance on the path of interest rates, quantitative easing (QE) should now be seen as a standard monetary policy tool. It has been used by multiple central banks around the world with little to no adverse consequences on their perceived independence.
Indeed, had the Bank not expanded QE through 2020, UK economic outcomes would not only have been materially worse and financial market functioning impaired, confidence in the Bank’s ability to meet its objectives would also have been greatly damaged. Ultimately, a central bank’s credibility comes from it doing the right thing to meet its mandate over the medium to long run. Expanding QE in 2020 was in our view the right thing to do in the circumstances and actually helped to boost its credibility and demonstrate its independence.
On the whole the Bank’s guidance on QE and other policy choices it faces around the lower bound for interest rates has been transparent. The Bank has given ongoing guidance on where it believes the effective lower bound to be and explained how its thinking has changed over time. Moreover, it has continued to update market participants around its expectation for how the unwind of QE would be sequenced in terms of whether rate increases or balance sheet reduction occurs first. This is all helpful from our perspective as a major market participant.
In Mark Carney’s “Lamda” Speech, the Bank provided a clear framework for how it views the trade-off between inflation and other economic outcomes. This helped to provide credibility for its flexible inflation target and allow it to ease policy after the Brexit referendum without unanchoring inflation expectations.
However, we believe there are some areas where communication could be improved:
- We would support clearer state contingency in the forward guidance around its asset purchase programme, so that it explicitly tied the run rate of QE (and the ultimate size of the balance sheet) to specific macroeconomic outcomes. We believe this can be a reasonably effective tool to provide further monetary easing.
- Policymakers could be more explicit about which channels they expect QE to be working through at any given time, given that QE might work in slightly different ways depending on the prevailing macroeconomic outlook (see answer to Q5).
- The Bank could provide more access to its internal macroeconomic model of the UK economy – COMPASS – and other propriety data it currently receives. On that note, we welcome the recent decision to publish the CHAPS data publicly through the ONS.
The experience of the financial crisis and its aftermath has increased the interest in changing central bank mandates across the world. The UK has not experienced the same persistent disinflation that many other advanced economies suffered following the financial crisis. The Bank’s flexible inflation target gives it the ability to look through temporary inflationary shocks, and adequately disentangle the impacts of supply and demand side shocks, as it did after the Brexit referendum led to a short-run inflation shock. Therefore, questions around the Bank’s mandate have been less pressing than in other jurisdictions.
However, we would welcome a more systematic review of the Bank’s mandate in line with the review the US Federal Reserve undertook through 2019 and 2020. In our view there are several issues in need of consideration:
- The secular decline in the equilibrium level of interest rates (the interest rate consistent with full-employment and inflation at target) has significantly increased the risk that any given negative shock will push interest rates to the effective lower bound, and so make supporting the economy more difficult.
- A standard inflation target does not contain any make-up provisions, so that past deviations from target are not corrected in the future. Were shocks to inflation evenly distributed, so that years of inflation above target were broadly offset by years of inflation below target, this wouldn’t be an issue as over time the price level would end up relatively close to target consistent levels. However, the skew of risks has clearly shifted towards more disinflationary outcomes, with many jurisdictions experiencing years of sustained below-target inflation. These inflation misses compound over time, leaving the price level a long way from the level economic participants reasonably expected.
Potential mandate shifts to deal with these issues include: a higher inflation target, a price level target, a nominal GDP target, and a wage growth target. Both a price level target and a nominal GDP target have desirable make-up properties, and would probably represent an improvement on the Bank’s current flexible inflation target.
QE is the large scale purchase of longer dated government bonds and other assets deemed appropriate by an independent central bank to meet its objectives. Crucially, these policies take place within, and are justified by, the central bank’s price stability objectives.
It is the use of these policies within an independent, transparent price stability framework which distinguishes QE from “monetisation”. QE might have the ostensible appearance of monetisation – gilts being purchased with newly created reserves – but the macroeconomic implications of this apparently similar operation differ significantly depending on the monetary framework under which they occur. Monetisation is done explicitly to finance government deficits with little to no consideration of price stability objectives. QE is sized and delivers in accordance with the need to provide monetary support to the economy to achieve price stability.
QE is typically thought of as a tool to provide extra stimulus when the economy is at the effective lower bound, when “normal” monetary policy is rendered much less effective. However, we find the distinction between “conventional” and “unconventional” or “normal” and “abnormal” monetary policy an increasingly unhelpful way to frame QE. Central banks have used multiple different tools throughout their history depending on the prevailing macroeconomic environment and whatever objectives they were trying to achieve. There is no reason to privilege the particular institutional arrangements that existed between roughly 1997 and 2008 in the UK, with bank rate the tool used to achieve an inflation target, as “normal” and everything else as an aberration.
Given changes to the macroeconomic environment, which are likely to leave the central bank rate much lower than it has been for much of history, QE is likely to be a recurring feature of monetary policy. It is firmly part of the monetary toolkit, and will be more appropriate in response to some future economic shocks, and less appropriate in response to others.
The overriding objective of QE has not changed throughout its life. Put simply, the objective has always been to provide additional effective loosening of financial conditions once the policy rate has reached the effective lower bound to help the central bank achieve its mandated objectives.
However, there are multiple channels through which QE can ease financial conditions, which at different points through the life of the programme have been more or less important. Furthermore, different policymakers have stressed different channels at different times (possibly reflecting their own view of how the policy works) which may have added to public confusion about the policy. To illustrate this point, we can distinguish between at least three channels through which QE works.
A liquidity effect, where the purchase of bonds by the central bank helps to smooth market functioning and provides financial institutions with the high quality liquid assets – i.e. bank reserves - they want. This channel only really operates in times of acute market stress, for example in the winter of 2008/2009 following the financial crisis and in March/April 2020 following the “dash for cash” in the early stages of the pandemic. In both of these cases, central bank purchases were vital for ensuring orderly markets, and helped ease the strains of acute liquidity crunches (which could have easily turned into solvency problems for various financial institutions). This channel can be thought of as working through the liability side of the Bank’s balance sheet, in that the tool gains traction through the provision of central bank liabilities – reserves – to other parties. This is the aspect of QE that is best described as something like “expanding the money supply”. However, outside of moments of market dysfunction, this liquidity or money supply expanding aspect of QE has a negligible impact on financial conditions.
There is also a portfolio rebalancing channel, which helps to push down long run interest rates by reducing the so called “term premium” on longer dated bonds. The idea is that some investors have quite particular preferences about the part of the yield curve they occupy. For example, regulatory pressure might mean that pension schemes prefer to hold longer dated gilts. These preferences mean that discrepancies in the price of different bonds along the yield curve, which we might normally expect to be arbitraged away, are persistent over time. These persistent barriers to arbitrage give central bank asset purchases the power to shift prices at the margin, putting downward pressure on long term rates, which in turn helps to boost other asset prices. This aspect of QE can therefore be thought of as working through the asset side of the central bank’s balance sheet in that it gets traction through buying particular assets. Indeed, this channel is potentially strongest when the Bank is purchasing non-gilt assets, such as corporate bonds, as this is where the barriers to arbitrage might be greatest. Because term premia is a difficult concept to measure, it can be difficult to determine the size of this channel, but generally speaking our view is that its size is overstated.
Instead, we tend to think the most important channel through which asset purchases impact financial conditions comes through a signalling channel, where QE alters the expectations of economic participants (see Vlieghe (2018) for a good summary of the evidence on this question[1]). This channel has multiple aspects. Firstly, in the way in which asset purchases reinforce the expectation of market participants that short term interest rates are going to be extremely low for a very long time. Given that long run interest rates reflect expectations about short term interest rates plus the term premium discussed above, changing expectations about the path of short rates mechanically changes the level of long run interest rates. By making investors expect short term rates to remain low for a long time, long term rates are also lower. Secondly, provision of QE may help to reassure investors that the central bank will do whatever it takes to stop extremely adverse shocks hitting the economy. Given that asset prices reflect the full distribution of risks, signalling a reduced risk of adverse tail risks helps to push up the value of riskier assets including shares and corporate bonds. Finally, to the extent that conducting QE boosts the credibility of the Bank’s inflation target, this should push up inflation expectations, reducing real rates, which in turn eases financial conditions.
Identifying the impact of QE is complicated by the multiple other confounding shocks which have also effected the economy since the policy was implemented, and indeed the feedback between the state of the economy and the degree of monetary support. These factors have made the identification of the effects of all monetary policy (i.e. not just QE) a difficult exercise in econometric analysis.
For example, one might look at the lacklustre performance of the UK economy following the financial crisis, and conclude that QE has done little to assist the Bank achieve its mandate and help the economy. However, this would be to ignore various headwinds the economy faces at this time, not least contractionary fiscal policy at a time when interest rates were at the effective lower bound. Indeed, in general, monetary policy is much more powerful at the effective lower bound when working in unison with expansive fiscal policy.
To get a clearer picture of the impact of QE it is necessary to consider how the economy would have behaved without it. To conduct this analysis, it is helpful to split the causal impact of QE on the economy into two steps. First, the impact that QE has on asset prices and financial conditions more generally. Second, the impact of this change in financial conditions on macroeconomic outcomes like growth and inflation.
Under certain textbook assumptions about the functioning of asset prices, the full impact of any change in monetary policy (including QE), should be felt in asset prices from the moment it is known to markets. Therefore, various attempts to study the impact of QE on asset prices have made use of event studies, which look at change in financial conditions immediately around a QE announcement. However, these studies face multiple confounding issues which complicate their findings, including the issue of disentangling the impact of other shocks affecting financial markets at the same time as the QE announcement, and indeed working out quite what constitutes an “announcement” given that many QE decisions were foreshadowed by policy makers. Moreover, these event studies do not capture the possibility that QE might have a more gradual impact on asset prices, perhaps because expectations take time to adjust or financial frictions impede immediate price adjustments. Finally, as discussed in response to question five, QE works through different channels which may be more or less powerful at any given time, which means the impact on financial markets may itself be contingent on the state of financial markets.
Given these considerations, it is not surprising that there is a wide interval of estimates of the impact of QE on financial markets. Joyce et al (2011) estimate that the Bank of England’s first wave of asset purchases from March 2009 to January 2010, which involved purchasing a cumulative £200 billion of gilts, led to an average fall in 5 to 25-year gilt yields of about 100 basis points[2]. Meaning and Warren (2015) estimate that the UK’s first £375 billion of QE reduced yields by around 25bps[3].Overall, the evidence of the UK’s case, combined with estimates of other QE episodes in other countries, points to a meaningful although not huge impact of QE on long term interest rates (see Haldane et al (2016) for a comprehensive summary of the evidence[4]).
Having established the impact of QE on asset prices it is relatively easy in principle to measure the impact of changes in financial conditions on the economy. Indeed we can measure this effect entirely separately from QE itself. All we need to do is tease out the relationship between the level and change in financial conditions and its impact on the economy.
Once we have this relationship, we can map the particular change in financial conditions that QE brought about with a particular impact it had on the economy. Using this method, Weale and Wieladek (2016) found that asset purchases worth 1% of GDP tend to boost GDP by 0.25% and inflation by 0.32%[5]. It should be noted that this analysis is much better at teasing out the impact of QE during periods of normal functioning. In periods where the liquidity aspect of QE (discussed above) was predominant, the relationship between QE and the economy is much harder to quantify as the effects are likely to be radically non-linear.
If the liquidity provision from QE kept markets functioning and certain institutions solvent – as the evidence suggests - this would have had hugely positive impacts on the economy compared to the counterfactual where markets were allowed to cease entirely and banks default. Indeed, the impact of QE is likely to change over time in part depending on the stance of fiscal policy. One reason why the round of QE started in the spring of last year may be more powerful than previous rounds is that fiscal policy is supportive. Fiscal and monetary coordination makes monetary and fiscal policy more effective.
QE, like all monetary policy, works through altering financial conditions. While easier financial conditions should not be conflated with greater financial instability, it is true that easier financial conditions in part gain traction in the economy by encouraging risk taking activity. Therefore, one potential trade-off of easing monetary policy is that it does increase financial instability through the build-up imbalances in various sectors of the economy.
However, monetary policy – including calibrating the level of QE - is unlikely to be the best tool to counter these risks; it is a rather blunt instrument in terms of targeting particular sectors where imbalances might be developing. It is precisely this bluntness that makes monetary policy such a powerful macroeconomic tool – it is able to affect the entire economy. But by the same token, it would therefore be inappropriate to keep the monetary pool too tight for the rest of the economy in order to tackle quite specific sector imbalances. Instead, macroprudential policy is a much more effective policy lever to tackle any incipient imbalances as it can be targeted directly to concerning sectors of the economy. Given the Bank’s current macroprudential tools and objectives, this trade-off is therefore unlikely to be binding.
To the extent that QE helps to ease financial conditions and therefore boost activity, this should help to increase inflation relative to what it otherwise would have been. Indeed, higher inflation is a feature not a bug of the programme. As discussed in other answers, the risk of this turning into runaway inflation is extremely low given the Bank’s clear price stability mandate.
QE largely works by easing financial conditions through higher asset prices. Unsurprisingly, those who own assets will therefore benefit from the policy. To the extent that asset ownership is inequitably distributed, then undoubtedly this will increase wealth inequality. However, we do not believe that QE in and of itself raises any especially problematic distribution issues for the following reasons:
All monetary policy – “conventional” or otherwise - has distributional consequences. Increasing or decreasing bank rate directly impacts the funding rate in interbank markets. This in turn impacts the price of a whole host of other assets throughout the economy. Moreover, given that a number of contracts are indexed directly to the level of bank rate, shifting bank rate alters the pattern of cash flows through the economy either to the benefit of lenders or borrowers. There is nothing unique, therefore, about the distributional consequences of QE compared to other monetary policy tools.
Simply looking at the impact of policy on asset prices provides an incomplete picture of the true distributional impact of QE. As discussed, by easing financial conditions, QE tends to put upward pressure on growth, wage growth and inflation and downward pressure on unemployment. In and of itself, reducing unemployment tends to have positive distributional consequences. But given that we know that downturns especially negatively affect the poor in terms of wage growth and employment, this impact is likely to be especially beneficial in terms of reducing inequality.
There are multiple tools that the government possesses to remedy any inequalities in the distribution of wealth and income. Issues of inequality are not the ambit of monetary policy, and indeed it would represent a deterioration of democratic accountability were they to be placed in the hands of monetary policymakers.
As discussed in our answer to Q7, there are two plausible channels through which asset purchases could undermine economic stability: increasing the risk of financial imbalances and increasing the risk of inflation expectations becoming unanchored. As we have defined QE – large scale asset purchases conducted within an appropriate, credible framework – we believe the risks stemming from either of these channels is small. Indeed, the more pressing risk to economic stability is that the Bank finds itself incapable of providing sufficient monetary stimulus to the economy to achieve its objectives.
Asset purchases outside such a framework are more problematic. For example, if the Bank was to purchase gilts well beyond the scope required to achieve its mandate it could be perceived as having subordinated its monetary policy objectives to government financing objectives and present significant risks to price stability. A robust macroprudential regime provides the Bank with the tools to ensure financial imbalances do not grow so large that they threaten financial stability.
There is no evidence that we are close to, or even that there exist, binding limits to the size of the Bank of England’s balance sheet. The size of the balance sheet is not a reliable indicator of policy or the likelihood of any particular macroeconomic outcomes. The risks and trade-offs associated with QE we have discussed in previous answers – financial imbalances and price stability – turn not on the size of QE purchases, but the justification for any asset purchases and the institutional framework under which they occur.
A better way of framing any potential limits to QE is to think about how the channels through which it works may become impaired in some way. As discussed in several other questions, there are multiple different channels through which QE operates which may be more or less powerful at any given time depending on the state of the economy. When financial market functioning is deeply impaired and banks are desperate for liquid assets, QE has a very powerful liquidity effect. However during periods of more normal market functioning, this effect is greatly attenuated making QE a less powerful tool.
The signalling and portfolio rebalancing channel both help to reduce long term interest rates by reducing the expectation of future rates and reducing term premia respectively. When the yield curve is extremely flat, with long term interest rates very low and not much higher than the policy rate (either because of rate expectations already being very low, or lack of term premium), then these channels are likely to be a lot less powerful. It is just much harder to push long rates much lower when they are already extremely low. That said, while QE may not be able to push rates much lower in these circumstances, they may still be an effective tool at reducing the risk of any future increase in rates; so they stop a counterfactual increase in rates that would have otherwise occurred were it not for the policy.
Of course these limitations speak only to the limitations of buying gilts as part of the QE programme. The Bank could purchase other riskier assets – for example corporate credit or even equity indices – which all have a risk premium over long term interest rates. QE purchases of these assets would likely have a powerful portfolio rebalancing effect, reducing the risk premium and increasing the price. This would help to ease financial conditions even if the long run risk free rate is close to zero and so boost the economy.
Finally, it is worth stressing that limits to QE, such as they are, are not limits to monetary policy per se. As discussed in Q12, a central bank could engage in ‘helicopter money’ where it credibly commits to increasing its liabilities without increasing its assets until certain macroeconomic objectives are achieved. It is close to impossible to see how helicopter money would not increase aggregate demand and so ultimately inflation. Ultimately, in a fiat money system, the central bank can set whatever price level it wants over the medium to long run.
It is very likely that the Bank’s balance sheet remains permanently larger than it was before QE was implemented. This reflects a structural increase in demand for the liabilities the bank issues – reserves and currency – due to regulatory pressure and changing business models encouraging banks to hold more of their assets in the form of high quality liquid assets.
The balance sheet is likely to shrink from its current size both in absolute terms and as a percentage of GDP. This will occur in part through allowing maturing assets to roll off the balance sheet and replacing them with new purchases; in part by active selling; and in part through economic growth. There is no reason to think this will provide any issues with regard to market functioning, and, if managed and signalled properly, the sale of gilts will probably exert only very little upward pressure on long run interest rates.
We foresee no issues involved in setting bank rate and exerting control and influence over various short term funding rates with a permanently larger balance sheet.
It is plausible the Bank could effectively write-off some of the government debt it has accumulated through its various QE programmes. In this case, the asset side of the balance sheet will shrink while the liability side is unchanged. This balance sheet operation is often called helicopter money, and has generated significant commentary recently. Once again, the operational framework under which this operation is conducted is crucial. If the Government directed the central bank to this action, then this would pose risks to price stability as it is effectively a demonstration of fiscal dominance. However, helicopter money, conducted by an independent central bank to achieve its price stability mandates, is capable of improving economic outcomes in periods of sustained disinflation. This policy move does not look likely in the current environment.
About us
Aberdeen Standard Investments (ASI) is a leading asset management company, formed in 2017 through the merger of Aberdeen Asset Management and Standard Life Investments. We currently manage £455.6bn[6] of assets. We operate ethically, encouraging good practices among companies we invest in and providing support and expertise for the long-term benefit of the communities in which we operate.
26 February 2021
[1] https://www.bankofengland.co.uk/-/media/boe/files/speech/2018/the-yield-curve-and-qe-speech-by-gertjan-vlieghe.pdf?la=en&hash=B7E9AF612B5DB7EEBAADEBCBD0A1EAA87FB1CF90
[2] https://www.ijcb.org/journal/ijcb11q3a5.pdf
[3] https://journals.sagepub.com/doi/abs/10.1177/002795011523400105
[4] https://www.bankofengland.co.uk/-/media/boe/files/working-paper/2016/qe-the-story-so-far.pdf?la=en&hash=8F7A0D4F0C0E466AACA9A03325776C2A13AAF55F
[5] https://www.sciencedirect.com/science/article/pii/S0304393216300101
[6] As at 30 June 2020