ECB Watch: The future of liquidity

The inflation shock of 2021/22 is forcing the ECB to think about policy normalization (and beyond) much earlier than anticipated by most central bankers. While the play book on interest rate policy is quite clear — hiking ECB key rates into restrictive territory and normalize once there is a strong conviction that inflation will revert to 2% — what to do with excess liquidity is less clear. Discussions on an update to the ECB's operating central bank liquidity framework are still ongoing. The task is planned to be concluded by year-end. Even now, however, we are confident to say that the ECB won't revert to the operating framework prior to the global financial crisis. Excess liquidity will stay with us! What level the ECB might judge as normal, how this level can best be achieved and what this means for interest rate markets is the topic of this note.

Let's start with some basics: what is excess liquidity? Excess liquidity is the liquidity within the euro area's banking sector, which finds no other purpose than being (voluntarily) deposited at the euro area's national central banks. Either within their current account in excess of minimum reserve requirements or in the deposit facility. Excess liquidity is, thus, calculated as the sum of the balance within the ECB's standing facilities (deposit facility - marginal lending facility) and the surplus within the banks' current accounts (current account holdings - minimum reserve requirements). Excess liquidity thus implies that minimum reserves are not binding and liquidity is not scarce (for the euro area as a whole). Read more on the ECB's current approach to minimum reserve policies in our recent note: ECB Watch: Minimum reserves and costly excess liquidity

Excess liquidity has increased materially since the global financial crisis. Before, the ECB applied a scarce liquidity framework in which pre-determined amounts of liquidity were provided to euro area banks through open market operations, supply and demand determined the exact interest rate with the ECB providing a minimum bid rate. This operating framework in which just enough liquidity is provided by the central bank to meet minimum reserve requirements (+balance structural demand for liquidity from exogenous sources) is known as a corridor system. The ECB's deposit rate and marginal lending rate built a corridor for money market rates and as the ECB target balanced liquidity conditions (no shortage but also no excess) money market rates traded close to the main refinancing rate. Asset purchases programmes and (T)LTROs have caused excess liquidity to surge in the 2010s and pandemic related measures provided another boost. Excess liquidity currently stands at around EUR 3,600 bn. One (intended) effect of strongly increasing the supply of reserves and by that the level of excess liquidity has been to firmly anchor money market interest rates. While short-term lending rates between euro area banks fluctuated around the ECB's main refinancing rate before the global financial crisis, when excess liquidity was close to zero, money market interest rates are now steered by the ECB's deposit rate. This is known as a floor system.

Money market rates pushed to ECB deposit rate as excess liquidity grew

Excess liquidity = (holdings in deposit facility - holdings in marginal lending facility) + (current account holdings - minimum reserve requirements)

Source: Refinitiv, ECB, RBI/Raiffeisen Research

The question which arises, and which is currently discussed within the ECB, is whether a system of excess liquidity will reflect the "new normal". And, if that's the case, which level of excess liquidity might be appropriate in an environment which should be neutral to the business cycle and inflation? Some guidance can be taken from the US Federal Reserve which has agreed to operate within an ample-reserves regime. An ample-reserves regime represents the idea that changes in the supply of reserves do not result in large changes in money market interest rates. Further, the level of reserves is kept high enough to ensure a buffer for short-term fluctuations in reserves not in the control of the central bank. That the appropriate level of reserves within an ample-reserves regime is not always easy to determine has become evident in mid-September 2019, when reserves unexpectedly became scarce and money market rates started to deviate from the Fed's monetary policy rates. Back then, reserves stood at around 9% of banks' assets, while with the benefit of hindsight a ratio of 11% might have been the threshold between an ample- and scarce-reserves regime. Currently, the ratio of reserve over banks' assets is at 14% and US money market show no signs of stress.

For the ECB the current level of excess liquidity can without doubt be classified as ample. Relating the spread of short-term money market rates and the ECB's deposit rate to the level of excess liquidity clarifies this (see chart). There is a clear negative non-linear relationship! The negative relationship implies that with a growing level of excess liquidity the downward pressure on money market interest rates grows pushing interest rates closer to the deposit rate or even below. But also the non-linearity is important as the slope of the relationship softens with the level of excess liquidity. Put differently, every additional increase in excess liquidity reduces money market rates by a smaller amount. By simply looking at the chart and assuming that the ECB decides to continue operating within an ample-reserves regime, keeping excess liquidity within a span of EUR 1,000 to 1,500 bn might seem a reasonable choice, below which money market interest rates react more strongly to changes in the supply of reserves. Taking into account the growth of the banking system over time brings the EUR 1,400 to 1,900 bn in today's terms. Compared to today's level of excess reserves of EUR 3,600 bn, thus, even in an ample-reserves regime the ECB can continue its balance sheet rundown.

Reduction of excess liquidity still feasible even in an ample liquidity framework

Money Market Rates refers to EONIA / €STR+8.5bp; Dots represent daily averages for a respective month.

Source: Refinitiv, ECB, RBI/Raiffeisen Research

Let's assume (for now) that the ECB decides to continue operating in an ample-reserves regime. Then, with EUR 600 bn TLTROs maturing until year-end 2024 this gives a scope for Quantitative Tightening (QT) of at least EUR 1,000 bn. If the ECB sticks to its current plan — no reinvestments in APP but full reinvestments in PEPP — the ECB's bond portfolio would decline by approximately EUR 500 bn until year-end 2024. Stepping up QT in 2025 by ending reinvestments in PEPP is expected to bring the euro area's excess liquidity close to the target motived above by year-end 2025. While this scenario, which forms the current consensus, is a possibility, it reflects a rather gradual approach. And also one, in which QT is stepped up at a time when also rate cuts are expected to occur. Thus, we think there is a case to be made that PEPP reinvestments might be stopped earlier than currently signalled as part of normalizing the ECB's balance sheet within an ample-reserves regime.

ECB balance sheet normalization sequencing: First TLTROs, then APP, then PEPP

Dashed lines: June 2023 Survey of Monetary Analysts; vertical axis: EUR bn.

Source: Refinitiv, ECB, RBI/Raiffeisen Research

The cornerstones of an ample-reserves regime have been well defined and tested by the Federal Reserve, why we think the ECB will find its euro area specific variant of an ample-reserves regime. This being said, there are other options available. While Isabel Schnabel pointed out in a recent speech that there is no way back to the pre-2008 framework, she was very committed to highlight those other options. In principle the alternative to an ample-reserve regime with a large bond portfolio on the central bank's asset side is to provide regular collateralised lending operations. This is the Bank of England's (BoE) approach. In August 2022, the BoE announced a new facility, the Short Term Repo (STR) Facility. In a nutshell, commercial banks can borrow reserves in unlimited amounts for 7 days against collateral. You might ask: what's the difference to the ECB's main refinancing operations (MROs)? It's only the interest rate, which the BoE sets at the 'Bank Rate' - a deposit rate. The ECB's MROs, on the other hand, are priced at the main refinancing rate, which is 50 bp above the deposit rate. Thus, the BoE still operates in a floor-system (money market interest rates are tightly steered by the deposit rate) but the size of the balance sheet is potentially smaller than in a Fed-style ample-reserves system. The reason is that in the BoE's framework, the amount of reserves is driven by the banks' demand for reserves which is expected to inflict a smaller liquidity buffer on the central bank's balance sheet compared to the Fed's framework, where the size of reserves is set by the central bank which does not hold full information about banks' liquidity needs and preferences. At the moment also for the BoE reserves are ample, as the STR facility is only intended to gain importance in a more advanced phase of QT. Thus, there is not much difference at the moment but in the long-term the asset side of the BoE's balance sheet should reflect a different structure than the Fed's. The BoE's asset side will be a mixture of security holdings and refinancing operations, the latter being only a liquidity backstop in the US.

If the ECB follows the BoE's path, this would not be a novum. Refinancing operations have always played a vital role in the ECB's monetary policy mix and the ECB could simply stick to conducting MROs. It would mean that the fixed-rate full-allotment characteristics, which have been introduced in 2014, would become standard. If the ECB is afraid of money market interest rates becoming too volatile, the interest rate could either be set at the deposit rate, like the BoE, or slightly above it, like the Swedish Riksbank which provides collateralised credit at 10 bp above the deposit rate. We would not expect the current asymmetric interest rate corridor to be kept. Anyway, what is important is that money market rates will also be closely linked to the ECB's deposit rate but being priced still somewhat higher than today. Like in an ample-reserves system it's not the interbank competition for reserves which will bid up money market interest rates, if liquidity providing open market operations are offered sufficiently close to the deposit rate. The key difference is the size of the bond portfolio. Thus, the BoE's framework would allow the ECB's bond portfolio to gradually unwind towards a level which is needed to ensure a well-functioning and highly liquid euro area bond market. The reserves providing function of the bond portfolio would be a secondary objective. While we also see ECB's MROs to gain importance as QT advances, we would not expect MROs to become the dominant source of providing reserves to the euro area banking sector but rather see the ECB's bond portfolio to remain key. Given the ECB's track record with applying refinancing operations, the idea of including demand-driven aspects in determining the necessary supply of reserves, the ECB might indeed find appealing.

MROs played no role in the post-GFC policy mix: is it time for a comeback?

Selected assets of the ECB's balance sheet in % of total assets.

Source: Refinitiv, ECB, RBI/Raiffeisen Research

In our view, the ECB might stick to an ample-reserves regime for the aggregate euro area, thus, holding a bond portfolio in excess of what is needed for pure market functioning purposes. However, given the uneven distribution of excess liquidity among euro area countries / banks an element of demand-driven reserves might be part of the ECB's operating framework going forward. What we mean is that banks holding less liquidity than they prefer will have the option to draw liquidity from the ECB at a fixed-rate with full-allotment against eligible collateral. This would act as an alternative to (secured) interbank lending and thus caps short-term money market interest rates. Short-term money market interest rates would continue to trade below the main refinancing rate but the €STR might trade in closer proximity to the ECB's deposit rate.

Let's wrap up!

• The ECB is currently working on an updated operating framework with respect to central bank liquidity. The outcome will determine the long-term size and composition of the ECB's balance sheet with implications for money markets and (sovereign) bond markets.
• It is clear that there is no way back to the pre-financial-crisis operating framework of scarce central bank liquidity. The ECB will keep a tight grip on money markets also in the future, as (interbank) markets are more fragmented and the regulatory requirements changed.
• Among leading central banks there are two options available. The Federal Reserve's supply-drive ample-reserves regime and the Bank of England's demand-driven approach. While the former provides liquidity predominantly through the bond portfolio, the latter allows for a larger drawdown of the bond portfolio being compensated by an active use of refinancing operations.
• For the ECB, we think that main refinancing operations will once again play a more active role in providing liquidity. However, we do not see them to play the dominant role but rather provide marginal liquidity to deal with the uneven distribution of excess liquidity within the euro area. Liquidity at fixed-rate and full-allotment against eligible collateral (introduced in 2014) will become a permanent characteristic and the interest rate corridor (gap between main refinancing rate and deposit rate) might be narrowed.
• We believe the ECB's bond portfolio will continue to be the dominant source of central bank liquidity. Put differently, we think the ECB's bond portfolio will be kept at a level to provide ample liquidity for the euro area as a whole. Given the current level of excess liquidity of EUR 3,600 bn, we would still consider this to be ensured with a reduction in excess liquidity of around EUR 2,000 bn. Thus, the ample-reserve regime does not stand in the ECB's way to push forward its balance sheet wind-down through maturing TLTROs and bond holdings.
• In this setting short-term money market interest rates would continue to be firmly anchored and trade below the main refinancing rate. Thus, the ECB's deposit rate will remain the main anchor point. Still, short-term money market interest rates would trade at higher spreads to the deposit rate as excess liquidity is lowered.