Speaking points by Peter Praet, Member of the Executive Board of the ECB, at the ECB Workshop on “Money markets, monetary policy implementation and central bank balance sheets”, Frankfurt am Main, 6 November 2017
It is my pleasure to welcome you to this year’s edition of the ECB´s Workshop on “Money markets, monetary policy implementation and central bank balance sheets”.
I would like to make some remarks on the role that money markets, monetary policy implementation and central bank balance sheets have played in our recent monetary policy decisions.
The October monetary policy decision
At its October meeting, the Governing Council decided to extend its net asset purchases within the asset purchase programme (APP) at a monthly pace of €30bn for nine months until the end of September 2018, or beyond, if necessary.
The Governing Council re-iterated that it expects the key ECB policy rates to remain at their present levels for an extended period of time, and well past the horizon of our net asset purchases.
This decision was taken against the backdrop of the solid and broad-based economic expansion which the euro area continues to experience.
Our monetary policy measures have been effective and are reflected in recent economic developments. Deflationary risks have disappeared. Some measures of underlying inflation have ticked up over recent months, but have yet to show more convincing signs of a sustained upward trend. Overall, inflation developments, despite the solid growth, have remained subdued.
Importantly, the convergence of inflation towards our aim remains conditional on a substantial degree of monetary policy accommodation.
A “sustained adjustment in the path of inflation” is the principal condition that has guided, and will be guiding, the calibration of our asset purchase programme (APP).
An extension of the programme was warranted by the continued need for a substantial amount of monetary policy accommodation to ensure the desired inflation convergence.
When considering the appropriate calibration of the APP there were three important dimensions to consider: pace, horizon and optionality.
- As regards the first dimension, pace, the brighter economic prospects have increased our confidence in the gradual convergence of inflation towards our aim. This called for a lower pace of purchases.
- Second dimension, horizon: We have always emphasised that monetary policy needs to be persistent and patient for underlying inflation pressures to gradually build up. This speaks in favour of a sufficiently long horizon of additional net purchases. The longer horizon also anchors short-term interest rate expectations for a longer period, thereby reinforcing the Governing Council’s forward guidance on policy rates.
- The third dimension is optionality: Retaining the option to re-calibrate the APP if warranted is consistent with the forward guidance on the APP.
Key channels of the APP
There are two key channels through which the APP operates. The first is the extraction of duration risk, which propagates through portfolio rebalancing. The second key channel is the signalling channel on interest rates.
The duration risk channel and the signalling channel correspond to the two components of the long-term interest rates through which a central bank can seek to influence the level and shape of the yield curve: the expectations component and the term premium.
How does the APP influence the term premium? By accumulating a portfolio of long-duration assets, the central bank extracts duration risk from private hands, frees up risk bearing capacity in the markets, spurs a rebalancing of private portfolios towards the remaining securities, and thereby lowers term premia and yields across a range of financial assets. Duration extraction is thus the catalyst for the portfolio rebalancing channel, which is the main mechanism by which easing through quantitative interventions is then transmitted further through the economy.
Now, what matters in particular for term premia is the entire stock of duration which the central bank removes from the hands of price-sensitive investors by taking this stock of duration on its balance sheet. As a result, the continued monetary support from assets purchases is not only provided by the additional net asset purchases, but also by the sizeable stock of already acquired assets and the forthcoming reinvestments.
The second key channel is the signalling channel: the expectations component of long-term yields reflects market expectations of the future path of the policy-controlled short-term interest rates. In this respect the sequencing embodied in our forward guidance gives significant strength to the signalling channel.
The sequencing firmly anchors expectations for short term interest rates. The Governing Council expects the key policy rates “to remain at their present levels for an extended period of time, and well past the horizon of our net asset purchases”.
The role and function of money markets
For the signalling channel to operate it is essential that the monetary impulses are transmitted through money market rates. In this regard let me make a number of observations.
First, for monetary policy impulses to be transmitted through money market rates, it is important that money markets are not fragmented. Last time I addressed this workshop, which was in 2013, the backdrop provided by market conditions was quite different from what it is now. At that time, we observed elevated fragmentation in money markets. This related in part to sovereign risks which were prevalent at that time.
Since the sovereign debt crisis, the overall levels of fragmentation have receded sharply. The dispersion of risk-adjusted borrowing rates across countries is now relatively low.
Of course, rate-based indicators do not give direct evidence of the degree to which trading is taking place across borders. But an assessment of fragmentation based on trading quantities would be anyway rather elusive. The high level of aggregate excess liquidity has significantly curtailed the need for trading liquidity.
The fact that money market rates, adjusted for credit risks, have largely converged speaks in favour of a much reduced level of fragmentation and a more homogenous degree of access to liquidity across the euro area. This is also supported by the fact that the interest rates observed across countries are more correlated with aggregate excess liquidity levels rather than country-specific levels.
Unsecured vs. secured segments
Second, we have observed a sustained shift of money market activity from the unsecured to the secured segment. Activity in the unsecured segment has declined significantly over time, in the first place due to the heightened counterparty risk in the wake of the crisis. Activity in the unsecured market has stabilised but remains subdued. This is linked to the high level of excess liquidity and reduced short-term funding needs for banks, as well as, to some extent, regulatory changes affecting banks’ liquidity management.
The ECB has never singled out one specific short-term interest rate as its operational target, but it is clear that the EONIA, an unsecured money market rate, has provided an important point of reference to practitioners.
The observed shift of activity from the unsecured to the secured market raises a number of important questions. For instance: Do unsecured rates still have the same information content as before the crisis? Is the arbitrage between secured and unsecured rates working efficiently? And, should the central bank focus more on steering secured rather than unsecured rates?
In this regard it is also relevant that we have seen some divergence between secured and unsecured rates, which also relates to some extent to our monetary policy measures.
While the ECB deposit facility rate has provided a floor for the EONIA, certain repo rates have come to trade significantly below OIS rates, especially for repos secured by collateral of the highest credit quality.
This reflects to some extent our monetary policy measures. By removing securities from the market, the ensuing scarcity exerts downward pressure on repo rates. For instance, specialness premia tend to be higher in those market segments where the Eurosystem has purchased more bonds relative to outstanding amounts.
Differentiation within the secured segments
Even looking at the unsecured segment, we see some differentiation of transaction rates, which results to some extent from the way the liquidity created by our measures interacts with market participants’ access to our deposit facility.
While unsecured trading between banks with access to our deposit facility is taking place above the deposit facility rate, institutions without access are willing to lend at rates below the deposit facility rate. The distribution of liquidity to entities without access to the deposit facility is amplified by our asset purchases, which mechanically channel liquidity to large financial institutions located outside of the euro area.
The phenomenon that we see of money market rates below policy rates is not unique to the euro area, but is visible in most, if not all, cases where asset purchases have significantly ramped up excess liquidity. Take for example the US, where the Federal Funds rate is trading below the Fed’s Interest Rate on Excess Reserves. This is due to the fact that the Fed Funds Rate to a significant extent reflects transactions between banks that can earn interest on reserves and Federal Home Loan Banks that cannot. Moreover, as US repo rates traded at even lower levels to the extent that they largely reflect trades with non-banks that do not hold a Fed account such as government sponsored entities and money market funds, the Fed introduced its overnight reverse repurchase agreement operations (RRPs), which has allowed it to effectively steer money market rates and engineer a successful lift-off of rates.
In the euro area, institutions without access to the deposit facility currently lend in money markets at rates below the deposit facility rate to those that have access to the facility. These low rates can be seen as reflecting the continued very accommodative stance of monetary policy. This situation, however, is worth studying further, as it is relevant when thinking about the operational framework in the longer run.
The brighter economic prospects have increased our confidence in the gradual convergence of inflation towards our aim. At the same time, a substantial amount of monetary accommodation continues to be necessary to secure the gradual convergence of inflation towards our inflation aim.
The continued monetary support is provided by the additional net asset purchases, by the sizeable stock of acquired assets and the forthcoming reinvestments, and by our forward guidance on interest rates.
To this end, an effective transmission of our monetary support through the money market remains essential. I thus look forward to hearing back about the discussions which you will be having about money markets. I think it is very useful and important that this workshop brings together academics, central bankers and market practitioners, which provides for a variety of perspectives. I wish you a very productive workshop.