They Said It: Recent ECB Policymaker Comments on Quantitative Tightening

17 November 2022

By David Barwick – FRANKFURT (Econostream) – We have assembled below a reasonably complete compendium of all recent public comments by European Central Bank Governing Council members touching on the subject of quantitative tightening, favouring direct quotes where possible.

By ‘recent’, we mean for the most part since the October 27 monetary policy meeting, though in cases where a given policymaker shared relevant views previously in particular detail, we include those comments.

Whilst leaving it to the interested reader to draw conclusions, we note that despite a certain and probably expected hawk/dove divide when it comes to enthusiasm for tackling QT, monetary authorities in general advocate approaching QT with care.

That includes not only the doves one can imagine might be willing to continue buying assets ad infinitum, but hawks as well, whether Mārtiņš Kazāks (‘We should not delay the start of QT too much, but at the same time we should be cautious, because we want the transmission mechanism to work.), Pierre Wunsch (‘Honestly, there is no reason why we should keep buying sovereign bonds, but at the same time we don’t have a lot of information of what it would mean to stop QE, and I would be for [being] cautious there.’) or Robert Holzmann (‘We are careful because the financial markets are to a greater extent uncertain than maybe a half a year or a year ago.’).

In view of the increased financial stability risks, confirmed again yesterday by the ECB’s latest Financial Stability Review, however, the emphasis on prudence need not surprise.


Christine Lagarde (ECB)

04 November 2022

‘While it is true that interest rates remain the most effective tool for addressing the high inflation environment, it is important to signal, as we continuously do in our monetary policy statement, that “we stand ready to adjust all of our instruments within our mandate to ensure that inflation returns to our medium-term inflation target”. In this context, as I explained at last week’s press conference, in December we will lay out the key principles for reducing the bond holdings purchased as part of our monetary policy portfolios.’

27 October 2022

‘Well, let us call it the reduction of our APP monetary portfolio. This is a matter that we have discussed at our last retreat amongst ourselves, the governors. We did not discuss the substantive issues today – deliberately – because we decided on a lot of issues. But what we decided is that we would pursue that discussion and we would decide the key principles of the reduction of our APP monetary portfolio in December. So that gives you a bit of an indication of when those key principles will be discussed and decided, and I will be very pleased to inform you about those principles at our next monetary policy meeting in December. And that has to be of course in advance of the decision to implement and to roll out this reduction.’


Frank Elderson (ECB)

15 November 2022

‘While policy rates remain our main instrument for steering monetary policy, we have made clear that we are ready to adjust all our instruments to maintain the consistency we need in our monetary policy to achieve our primary objective of price stability.’


Luis de Guindos (ECB)

14 November 2022

‘The policy decisions we will take at our next meeting will be based on various elements, including our December macroeconomic projections. At this meeting, we also expect to lay out the key principles for reducing the bond holdings in our monetary policy portfolios. We will proceed with prudence, continuing to normalise our monetary policy in line with our medium-term price stability objective.’

08 November 2022

‘We will start with [QT] sooner or later, for sure in 2023. This process will have two positive effects: it will reduce excess liquidity and alleviate collateral scarcity. But quantitative tightening (QT) must be implemented with a lot of prudence. In my view, we should start with a passive QT by not fully reinvesting the maturing securities in our portfolio. The characteristics and the timing of our QT, which may overlap or not with the process of normalising the interest rates, will be discussed in December. Personally, I don’t see any sort of sequencing here. The first instrument that we have used, because we believe it to be the most efficient one, is interest rates. QT is also part of the normalisation process of monetary policy and we will proceed with a lot of prudence and caution.’


Fabio Panetta (ECB)

14 November 2022

‘…we should not ignore the fact that the tightening, which has followed from our decisions since the end of 2021 and from expectations of further adjustments in our stance, is already working its way through the economy – with the usual transmission lags. Estimates suggest that this tightening will on average subtract more than one percentage point from annual real GDP growth each year until 2024 compared to a counterfactual where interest rates and balance sheet expectations would have remained unchanged since December 2021.’

03 November 2022

‘…we must be clear about the sequencing of the normalisation process. We should avoid “cliff effects”, continually monitor the market response to our measures and consider the feedback between our different instruments. Currently, our policy rate remains a suitable marginal instrument of normalisation. It is the instrument we know best. We have a comparatively limited understanding of the effects of reducing the size of our balance sheet. The size of our balance sheet will be significantly reduced as targeted longer-term refinancing operations (TLTROs) mature and banks likely make early repayments after the decision we took last week to adapt the TLTROs’ terms and conditions to the current monetary policy context. We should take the necessary time to assess the impact of our rate hikes and of phasing out the TLTROs. As we normalise our monetary policy, we should expect bank lending conditions to tighten. What we need to avoid, though, is a sudden stop in the supply of credit to the broad economy. We should ensure that TLTRO repayments have been absorbed before we stop fully reinvesting the principal payments from maturing securities purchased under our purchase programmes. And when considering how we would then reduce the size of our bond portfolios, a controlled reduction – whereby only redemptions above a cap are not rolled over – is preferable to active sales, which may unsettle markets in an already volatile financial environment.’


Pablo Hernández de Cos (Banco de España)

16 November 2022

‘In my view, it may be appropriate to wait until sufficient outstanding TLTRO III balances have been repaid before starting the bond portfolio run-off, for two reasons. First, the large TLTRO repayments expected in the first part of 2023 will give us an initial picture of any potential asymmetry and non-linearities in the effects of balance sheet reduction vis-à-vis those generated by balance sheet expansion, by when we will probably have too a clearer idea about whether we are in, or headed towards, a recession. Second, it seems reasonable to wait and observe potential ripple effects of TLTRO III repayments in bond markets before starting the bond portfolio run-off, as it will allow us to assess the effective impact of balance sheet reduction on financial conditions. After first allowing markets to absorb significant TLTRO III repayments, the normalisation of the ECB’s balance sheet could be followed by ending full reinvestment of the APP portfolio. As announced by President Lagarde last month, we will decide on the key principles of the reduction of our APP portfolio at our December monetary policy meeting. And why we should end APP reinvestments before PEPP reinvestments? In my opinion, a “first-in-first-out” staggered approach is preferable for at least two reasons. First, PEPP reinvestments remain an important line of defence against (pandemic-related) fragmentation episodes, so they are potentially pledged for reallocation across jurisdictions while the lasting vulnerabilities caused by the pandemic continue to pose a risk to the smooth transmission of our monetary policy. Second, we previously declared that PEPP reinvestments will not end up until at least the end of 2024, while our forward guidance on APP reinvestments was open-ended. We have clearly stated our intention to continue APP reinvestments for an extended period of time past the date when we started raising the key ECB interest rates. Moreover, our balance sheet policy will depend too on the operational framework that we adopt for the foreseeable future. Since the financial crisis, we have been de facto operating a “floor” system, characterised by ample excess reserves and interbank rates pegged to the DFR. This floor system, similar to that adopted by the Federal Reserve, is different from the “corridor” system that we operated before the crisis, which was characterised by a leaner balance sheet and scarcity of reserves. The decision we take in this respect, as well as regarding the desired long-run composition of assets and liabilities, will anchor the desired size of our balance sheet. Once we reach the point of reducing asset holdings acquired through the APP, the crucial question is how strong an effect this form of quantitative tightening will have on long-term interest rates. It is tempting to see quantitative tightening as simply the opposite of quantitative easing. In this light, our experiences with QE over the past decade and more may give us some idea as to how large the effects of QT may be. So far, however, there is simply not as much historical evidence available about the impact of QT, because we have observed more episodes of balance sheet expansion than of balance sheet contraction. A number of considerations support a prudent approach, since the response to QT could be stronger than the response to QE. First, the fiscal landscape has deteriorated in recent years. Public debt to output ratios have climbed sharply since the onset of the pandemic, because governments have run very large deficits over the course of the pandemic and because output has fallen. The current inflation surprise somewhat reduces the real value of this debt, but this effect is small compared with the strong increase in debt levels since the pandemic began. The greater the amount of debt that private markets must absorb, the lower the liquidity in sovereign debt markets and the higher the yields that markets will demand in order to equate supply with demand. Accordingly, this factor tends to magnify the impact of quantitative tightening on the term premium. Second, the impact of balance sheet policies may also vary depending on liquidity conditions. In QE times, more liquidity injected into the system did not have a major impact since the market was saturated with excess cash and thus money market conditions were rather insensitive to additional liquidity injections. But in QT times, the gradual withdrawal of excess liquidity will make money market conditions increasingly sensitive to the amount of cash that is withdrawn from the system, and this could mean more volatility in the overnight rate transmitting in the form of more term premium throughout the curve. Indeed, several episodes in the past have shown these potential effects associated with QT. There are, however, some arguments that point towards attenuated effects of QT compared to QE. First, markets are already pricing in QT to some extent. This is important because much of the impact of asset purchase programmes stems from the advance information provided to financial markets about their overall scale and design. Since markets are forward-looking, the whole expected future path of asset holdings plays a role in determining market yields at any given point in time. That is what is typically known as “stock” or “announcement” effects. In contrast, beyond potential “stock effects”, several studies that have sought to measure the “flow effects” of asset purchases – the changes in yields directly related to the amount purchased or held at a given point in time – tend to conclude that these effects are relatively smaller.13 For this reason, we might expect that yields already reflect, at least to certain extent, current expectations on the implementation of balance sheet reductions. 14 However, by the same token, a faster balance sheet reduction than currently anticipated by markets could trigger some yield tightening. Second, compared with QE, QT conveys much less information about the future path of interest rates. Purchases and interest rates were linked in the ECB’s forward guidance, as I have already explained, but the ECB no longer employs forward guidance on rates as a policy instrument. Therefore, the ECB’s balance sheet adjustments alone no longer provide direct information about the timing of its interest rate adjustments. Again, this weakens the announcement effects of QT compared with QE. In sum, all these arguments point, in my view, to the need for the balance sheet reduction in the euro area to be very gradual and predictable. It is also essential that balance sheet reductions leave room for action against fragmentation if it reappears, whether either through flexibility in PEPP reinvestments or through activation of the TPI should this prove justified and necessary.’

02 November 2022

‘At some point we’ll have to start reducing our APP portfolio. We will decide on its key principles in December. In my view we have to do it in a cautious and very gradual manner. Mainly because we do not have much evidence on its effects. We need first to draw lessons from the reduction of the balance sheet stemming from the decision to modify the conditions for the long-term loans granted to banks – the TLTROs. And the reduction of the stock of APP assets will have to be considered together with our rate path, so as to achieve the financial conditions compatible with achieving our inflation target.’

02 November 2022

‘… at some point we will also have to start the reduction of the stock of assets on our balance sheet. Indeed, we will decide on the key principles of this reduction as regards the APP at our December meeting. In my view, we have to be cautious with this reduction and do it in a very gradual way, given the limited evidence we have on its effects. I think we need to first draw lessons from the balance sheet reduction that will follow from our decision last Thursday to modify the conditions of the TLTROs. And, obviously, the reduction in the stock of APP assets has to be taken together with those related to our path of interest rate increases, so that we achieve the financial conditions that we believe are compatible with achieving our objective.’


Joachim Nagel (Bundesbank)

10 November 2022

‘The monetary policy change of course also includes tackling the reduction of the Eurosystem's bond holdings. They amount to almost €5 trillion. Maturing bonds are still being replaced again under both the PEPP and the APP. In the wake of rising key interest rates, the question increasingly arises as to why the reinvestment policy tends to slow down the development of bond yields in the euro area. In my opinion, it does not fit together to move interest rates at the short end of the market in one direction and to influence those for longer maturities in the other direction. When you have two instruments for normalising monetary policy, it doesn't make sense to use only one of them.’

02 November 2022

‘I think it is very important to have this discussion regarding our balance sheet. When acting against inflation, you have to think about the bond portfolio, too, which has increased massively over the past years. Now we are talking about nearly 5 trillion euro of bond holdings. We want to normalise monetary policy. And to achieve this, we should use all the instruments we have at our disposal. We should start reducing the size of our bond portfolio at the beginning of next year, for example by letting existing bonds mature with little market impact.’


François Villeroy de Galhau (Banque de France)

11 October 2022

‘Another issue that we will need to confront is the normalization of our balance sheet. It has expanded in recent years through numerous programmes (APP, PEPP, TLTROs,..) and is now just short of €9 trillion. It would not be consistent to keep a very large balance-sheet for too long in order to compress the term premium, whilst at the same time contemplating tightening policy rates above neutral. In addition, rising remuneration on very large excess reserves may alter the transmission of the desired tightening through the bank channel. But obviously, this question is less pressing than the rise of interest rates to neutral, and should come only at a later stage. Let me simply put forward at this stage a few preliminary principles that could in my personal view guide the normalisation of our balance sheet, in due time:

  • First, our key interest rate should remain our primary instrument to adjust our monetary policy stance. The size of our balance sheet should be used as a complementary policy tool, whose effects are more difficult to calibrate or fine-tune.
  • Second, we should follow a clear sequencing regarding the various programmes: (i) the reimbursement of TLTROs comes first, and we should avoid any unintended incentives to delay repayments by banks (ii) at the other end, PEPP should be reinvested until the end of 2024, as stated (iii) this leaves inbetween APP for which the Governing Council has said that reinvestment will continue in full until “well past the date when it started raising the key ECB interest rates”. Here we could start earlier than 2024, maintaining partial reinvestments but at a gradually reduced pace.
  • Third, the phasing out of the asset portfolio should be orderly, announced cautiously and well in advance. The ‘stock effect’ of asset holdings, which is the key transmission channel, primarily depends upon the expected end-point of the balance sheet normalisation, in terms of both the terminal date and size. Fluctuations in the pace of the run-off during the travel to destination do matter less.
  • Fourth, as taught by US experience in 2017-2019, the pace of the balance sheet reduction should not be left completely on “automatic pilot”. Starting slowly, assessing markets reaction, and gradually accelerating seems like a sound approach. Some flexibility should be kept, in case of sudden liquidity shocks. Indeed, the ‘flow effect’ may temporarily play a role when market liquidity abruptly dries up.’


Klaas Knot (De Nederlandsche Bank)

15 October 2022

‘The level to which the policy rate will be increased also depends on the calibration of other tools, including the run-off of the central bank balance sheet. Once we will have reached neutral territory with our policy rate, it makes sense to consider the roll-off of asset purchases by limiting reinvestments. This would help transmitting our tighter monetary policy evenly across the yield curve and ensure that all monetary instruments work in the same direction. Model simulations indicate that lifting the policy rate in tandem with an upward term premium effect helps to reach the target more efficiently, as it lowers both peak inflation and the required terminal rate. I do believe that we should move gradually here. To this end, it makes sense to distinguish APP from PEPP. The former is an instrument exclusively to steer the monetary stance and should be seen as a complement to the policy rate. PEPP plays a dual role, by also countering serious risks to monetary transmission related to the pandemic. It implies that decisions about unwinding these programmes do not have to run in parallel per se.’


Ignazio Visco (Banca d‘Italia)

31 October 2022 & 03 November 2022

‘This recalibration [of TLTRO III] aims to strengthen the pass-through of key interest rate increases to bank lending conditions and to remove deterrents to the Eurosystem’s balance sheet reduction process through the early repayment of TLTRO III operations. The ECB Governing Council postponed the debate on the timing and arrangements of a gradual review of the reinvestment of principal payments from maturing securities under the asset purchase programmes, while maintaining the flexibility associated with the reinvestment of those under the pandemic emergency purchase programme (PEPP) over the coming months.’


Mārtiņš Kazāks (Latvijas Banka)

07 November 2022

‘The policy accommodation of the APP isn’t appropriate anymore, and if a central bank market intervention isn’t needed, then we have to withdraw from the market.’

Discussion of assets acquired under the pandemic emergency purchase programme (PEPP), in contrast, ‘shouldn’t start yet – there’s no need’, he said. In general, however, the discussion that Lagarde indicated was underway with respect to QT will continue next month, he said.

While some decision pertaining to QT could thus emerge from the December 15 meeting, the communication of a precise sequence of steps wouldn’t necessarily be forthcoming, he indicated. ‘We know that forward guidance in times of uncertainty is not always helpful. But from today’s perspective, in my view it would be appropriate to introduce an element of balance sheet adjustment via the APP early next year unless the economic outlook changes sharply.’

Asked whether the time to actively sell parts of the ECB’s balance sheet could come in 2023, Kazāks said he ‘would not jump that far ahead, because uncertainty is very high’.

He continued: ‘We should not delay the start of QT too much, but at the same time we should be cautious, because we want the transmission mechanism to work. Later on, we should and will talk about actively selling our asset holdings. But it’s still way too early for that.’

Kazāks hesitated to back the view that the modification of the TLTRO conditions would ease pressure on the ECB to unwind its balance sheet via QT. ‘I would look at them as addressing two different segments’, he said. ‘But a reduction in TLTRO money of course would align our instruments more with the interest rates.’

Even with early repayments of the TLTROs and the consequent reduction in liquidity, it would be ‘perhaps years that we will continue to see excess liquidity in the system, given how much is there’, he said, leaving the deposit facility rate the relevant key interest rate for some time to come.

The idea of maintaining reinvestment of maturing securities at least until TLTRO repayments have been absorbed did not win Kazāks’ endorsement. ‘I would refrain from that specific sequencing of instruments’, he said.


Yannis Stournaras (Bank of Greece)

15 November 2022

‘We will actually discuss the prospects of future steps in this direction in our December Governing Council meeting. Any such steps should be cautious and gradual, as quantitative tightening reinforces interest rate increases across the yield curve. In fact, in a recent Bank of Greece Working Paper (“A global monetary policy factor in sovereign bond yields” by Dimitris Malliaropulos and Petros Migiakis, July 2022), the authors found that global central banks’ large-scale asset purchases have contributed to significant and persistent declines in long-term yields globally, ranging from around 330 basis points for AAA-rated sovereigns to 800 basis points for non-investment grade sovereigns. Hence, tightening monetary policy by scaling down central banks’ balance sheets to pre-crisis levels may lead to sharp increases in sovereign bond yields globally and widening spreads of vulnerable sovereigns, with severe consequences for financial stability and the economic prospects. In the euro area, we stand ready to use the Transmission Protection Instrument, to the extent needed, to counter unwarranted, disorderly market dynamics that pose a threat to the transmission of monetary policy across all member countries.’

04 November 2022

‘Any such steps should be cautious and gradual, as quantitative tightening reinforces interest rate increases across the yield curve.’

‘Scaling down central banks’ balance sheets to pre-crisis levels may lead to sharp increases in sovereign bond yields globally and widening spreads of vulnerable sovereigns, with severe consequences for financial stability and the economic prospects.’

It is the Governing Council’s consensus to only start reducing its bond holdings after the central bank is done with hiking rates.


Gabriel Makhlouf (Central Bank of Ireland)

16 November 2022

‘It might be tempting to assume that the effects of reducing the size of the balance sheet will mirror the effects from increasing it, but there are some key differences. The primary difference now is the context. The current high inflation is having a very negative impact. Falling real incomes mean lower consumer spending as households reduce spending on non-essentials, and divert more of their budget to items such as food and energy.’


Pierre Wunsch (National Bank of Belgium)

08 November 2022

‘Honestly, there is no reason why we should keep buying sovereign bonds, but at the same time we don’t have a lot of information of what it would mean to stop QE, and I would be for [being] cautious there.’

QT should ‘start early and cautiously’, with an assessment shortly into the exercise of whether ‘there’s enough absorption capacity in the market’ before picking up the pace.


Constantinos Herodotou (Central Bank of Cyprus)

14 November 2022

‘Of course, at a later stage and when deemed necessary, the reduction of the ECB's balance sheet can also be discussed. In any case, I believe that the best approach is to continue to determine any moves we make on the basis of the financial data that we will have in front of us at each meeting of the Board, without predetermined directions.’


Peter Kažimír (National Bank of Slovakia)

10 November 2022

The ECB’s ‘price stability mandate requires us to reduce our presence in the market’. This is ‘absolutely vital for our credibility.’


Boštjan Vasle (Bank of Slovenia)

10 November 2022

The ECB needs to unwind its unconventional instruments.


Robert Holzmann (Austrian National Bank)

28 September 2022

The ECB wants to reduce its balance sheet. ‘We are careful because the financial markets are to a greater extent uncertain than maybe a half a year or a year ago.’ The ECB does not wish to ‘overwhelm’ the markets. There are moreover ‘many, many technical questions, economic questions’ related to the reduction of the balance sheet. ‘We will be careful about reducing it.’