Peter Praet: Current issues in monetary policy
Speech by Mr Peter Praet, Member of the Executive Board of the European Central Bank, at the Peterson Institute for International Economics, Washington DC, 9 December 2014.
I would like to thank Tobias Blattner for his contribution to this speech.
The ECB has responded forcefully to counter growing risks of a too prolonged period of low inflation. The current set of measures will enhance our monetary policy stance, incentivise bank lending, overcome remaining impairments in our transmission mechanism and underpin the firm anchoring of medium- to longer-term inflation expectations. Tentative evidence suggests that our credit easing package is delivering some tangible first benefits. But we need to remain vigilant as to whether the force of monetary stimulus already in the pipeline remains sufficient to reach our objective. If needed, one option for the ECB to ease the monetary policy stance further would be to extend its asset purchases towards other asset classes. Purchases of different asset classes will affect private sector financing conditions to varying degrees as they will activate different transmission channels and affect different spread components.
* * *
Ladies and Gentlemen,
Let me thank the organisers for inviting me here today.
Konrad Adenauer - Germany's first post-war Chancellor - once famously said, (and freely translated): "Why should I care about remarks I made in the past?" He said so knowing that the speed at which events unfold may - at times -outpace our comprehension of them at any given moment. And when I went back to the remarks I gave here in the spring of last year, my impression was that the pace of events over the past year or so was indeed dramatic.
In May last year I discussed two things.
I discussed the challenges monetary policy faces when banks engage in rapid leveraging and deleveraging. I noted that "the central bank may have to be vigilant that the pace of asset reduction by banks is commensurate with an evolution of income, activity, and employment that does not create downside risks to price stability".
And I discussed the long-term strategy of the ECB. Here I highlighted two main attributes of this strategy, namely that "it must be sufficiently flexible to adjust its policy conduct to the specific challenges at hand. And it must be sufficiently binding to anchor expectations and build central bank credibility with market participants and the general public".
When reading through my notes, I asked myself: did we indeed have a steady eye on the state of the banking sector? And have we applied the two principles of our strategy - that it is flexible and binding - in what turned out to be a very challenging year or so for monetary policy in the euro area?
I think the answer to these questions is a qualified yes.
A flexible policy response
Let me start from the spring of this year.
Our diagnosis at that time was that borrowing rates were becoming too unreactive to policy changes. We saw that our policy impulses - a combined cut of 50bps in our key policy rate over the preceding year - were no longer being transmitted to private borrowing costs in the way they used to.
As a result - and despite our very accommodative stance - bank lending rates remained at elevated levels in large parts of the euro area. This meant that weak credit dynamics were increasingly becoming a source of downside pressures regarding our medium-term inflation outlook.
In other words, a dysfunctional bank lending channel was in full display: excessively restrictive borrowing conditions were suppressing demand in the countries where they were applied, and indirectly - through intra-euro area trade linkages - were contributing to economic weakness in the rest of the euro area.
So, we had a very close eye on the banking sector as you can see. How did we respond to our diagnosis?
In the face of these risks, we did precisely what any central bank would have done: we acted to restore transmission and bring down the average borrowing costs paid by households and firms to levels more consistent with the intended stance of our policy.
This was the philosophy underlying our credit easing package that we adopted between June and October this year.
The transmission mechanism we wanted to activate - the bank lending channel - was quite simple and, I would add, orthodox in its underlying logic.
Our targeted long-term refinancing operations (TLTROs) provide significant term funding relief for banks: a very low fixed rate over a period of up to four years. They are designed to maximise the chances that banks pass the funding relief on to their borrowers. Banks will have to repay loans in advance if they fail to meet the agreed lending benchmark. And the more they lend beyond the benchmark, the more they are allowed to borrow. This positive incentive is amplified by the use of a multiplier.
This shows clearly that we put incentives first, employing them to counteract the existing impairments in the bank lending channel.
The ABS purchase programme (ABSPP), which we first announced in June and which became operational in November, will help further lubricate the transmission process. It can activate the same scarcity channel that was found to be a critical ingredient in the success of the Fed's Large Scale Asset Purchase Programme.
This channel works as follows: ABS purchases push down the market spreads paid on senior tranches, currently the main focus of our purchases. They do so by subtracting volumes from potential private investors in such tranches, and thus encourage banks to relieve the resulting scarcity by originating more ABS. Ultimately, they can only do this by creating more loans.
The result is again an increase in loan supply. And as the supply of loans shifts to the right along a given demand for credit, lending rates adjust downward to create the necessary demand.
In September we added an additional leg to our June package: purchases of euro area covered bonds (CBPP3), again with a view to making credit more abundant and less expensive. Covered bonds share some of the characteristics of the ABS: notably, they are loan-backed securities where the same scarcity channel can be put to work.
Importantly, the CBPP3 has also strengthened the quantitative dimension of our policy. By increasing the pool of purchasable assets - and by communicating our intention that our combined measures will help bring the ECB's balance sheet back to levels it had in early 2012 - we are providing guidance on the overall scale of our interventions; a scale considered necessary to deliver the intended support to inflation and the recovery of the euro area economy.
And in scaling up the volume of our measures, we have also broadened the scope of transmission. The TLTROs and the ABSPP primarily work through what is called a direct price and pass-through effect. That is, both measures are designed to ensure an unimpeded transmission of our actions to the ultimate borrowing costs of the private sector.
Purchases in the covered bond market will have a similar effect, but because of the size and liquidity of the market, other transmission channels may become more important in relative terms. In particular, our covered bond purchases will also activate two prime transmission channels of large-scale asset purchases: the portfolio rebalancing and signalling channels, which are more dependent on the size of the monetary stimulus.
So, I think it is fair to say that our response so far has been flexible. The diagnosis that led to our credit easing package identified a dysfunctional bank lending channel as a prime threat to achieving our ultimate objective of price stability. Our measures were tailor-made to that objective and offer a powerful response to address the root cause of the impaired bank lending, and thus to facilitate new credit flows to the economy.
Is it working? Tentative evidence suggests yes: our credit easing package is indeed delivering some tangible first benefits.
Broad measures of nominal financial conditions have eased and spreads on senior ABS and covered bonds have fallen. Consistent with our expectations, price adjustments have been particularly pronounced in those segments where we intervened directly. The case of the ABS market stands out: spreads fell by an impressive 70bps since May although our purchases only started a few weeks ago. By way of comparison: the Fed's QE announcements, which involved combined purchases of nearly $2tn of MBS, Treasuries and GSEs, are estimated to have lowered 30-year MBS yields by some 100bps.
At the same time, the broad-based nature of the easing across markets suggests that the signalling effect - operating through a flattening of the OIS curve - has also been a powerful element of our response.
Lending rates have fallen too. Since May the composite indicator of the nominal cost of bank borrowing for euro area NFCs fell by 40bps to 2.6% in October. In Italy this indicator fell by nearly 60bps. In other words, the premium Italian firms pay for bank borrowing relative to the average euro area firm has fallen in recent months - a sign that the transmission of our policy is gaining traction.
Lending volumes seem to be responding too. Evidence is growing that a turning point in loan dynamics might have been reached. As a result, the downward pressure on the output gap stemming from impaired credit supply is levelling off. At the same time, banks also report an increase in loan demand, which means that our measures come at the right time: banks would not borrow under TLTRO unless demand for loans strengthens.
A binding policy response
But flexibility aside, have we also adhered to the other element of our strategy I mentioned last year: have our measures been binding enough? Here we need to be more cautious in our assessment.
The fact is that - despite our broad-based measures and the associated fall in private borrowing costs - inflation has not stopped trending down. In November inflation in the euro area fell back to a cyclical low of 0.3%. When we last met here in May 2013 it was still at 1.4%. The sharp fall in oil prices since the summer - down by 30% in euro terms since early June - explains a significant portion of this fall.
But measures of underlying inflation - stripping out volatile items such as energy and food - have also continued trending down. Falling core inflation points to weakness in broad aggregate demand - consistent with the loss of economic momentum we observed in recent months. And our latest staff projections foresee a material downgrade of the macroeconomic outlook.
These developments have also affected inflation expectations. Short-term inflation expectations have almost moved in tandem with the fall in headline inflation. But more recently - basically since the summer - we have also observed a fall in medium- to long-term inflation expectations. In principle, these should not be too reactive to short-term commodity price gyrations.
Does this mean our actions have not been binding enough?
Not necessarily. First, notwithstanding their recent decline, most market- and survey-based measures of medium-term inflation expectations remain in close vicinity to our aim to keep inflation close to, but below 2%. Second, the fall in market-based measures was not limited to the euro area. Break-even rates and medium-term inflation swaps also fell in other jurisdictions, such as in the US.
This of course does not mean that we can ignore these developments. Well-anchored inflation expectations are indispensable for medium-term price stability. They provide a nominal anchor for the economy. And even more so in the current environment: given the potency of the current oil price shock, the risk is that inflation may temporarily fall into negative territory in coming months.
Normally, any central bank would prefer to look through a positive supply shock. After all, lower oil prices boost real incomes and may lead to higher output in the future. But we may not have that luxury at present. The reason is that shocks can change: in certain circumstances supply shocks can morph into demand shocks via second-round effects. In these conditions monetary policy needs to react to what initially appeared to be a supply shock, so as to prevent a destabilisation of inflation expectations.
Remember: this is the binding element of our strategy.
In light of this, the Governing Council last week reiterated that should the monetary stimulus fall short of our intention to move our balance sheet towards the dimension it had at the beginning of 2012, or should it become necessary to further address risks of a too prolonged a period of low inflation, it remains unanimous in its commitment to using additional unconventional instruments within its mandate. This would imply altering early next year the size, pace and composition of our measures. ECB staff and the relevant Eurosystem committees have therefore stepped up the technical preparations for further measures with a view to implementing them in a timely manner, if needed. Indeed, all of our monetary policy measures are geared towards underpinning the firm anchoring of medium to long-term inflation expectations, in line with our aim of achieving inflation rates below, but close to, 2%, and contribute to a return of inflation rates towards that level.
Extending the policy response
Extending our outright asset purchases towards other asset classes could be one option if we were to judge that the economy was in need of further stimulus. It is important though to constantly keep in mind our intention, our purpose in adopting further measures. Asset purchases are not an end in themselves. They are not a target of monetary policy. They are an instrument of policy.
To see this, let us conduct a thought experiment: faced with the circumstances we have now, how would we have reacted in normal times - that is, with positive policy rates and no impairments in transmission?
In my view we would have lowered rates further by now. This reaction would have been consistent with and guided by the wealth of evidence gathered over time on the transmission of changes in our key interest rates to borrowing rates and to economic activity and inflation.
So, when we say our measures intend to bring our balance sheet back to levels seen in early 2012, what we are really saying is that - given the current composition of the unconventional monetary policies that we have put in place - this is the amount needed to bring inflation in the medium term back to levels closer to 2%.
And in a bank-based economy such as the euro area, pinning down this amount rests crucially on the effectiveness of our measures in reducing the borrowing costs of households and firms.
It follows from this that an important criterion for the choice of additional measures - if needed - should be their degree of influence over broad financing conditions in the private economy, scaled up to an amount of easing judged sufficient to achieve our price stability mandate.
This might give rise to a trade-off: some assets might ex ante be more effective than others in reducing private borrowing costs, but might be in short supply. Others might be in abundant supply, but less effective a priori and plagued by institutional complications.
Understanding the effectiveness - and hence the transmission - of purchases is therefore key to calibrating both the size and composition of our response in case we need to act further.
Purchases of different asset classes will affect private sector financing conditions to varying degrees as they will activate different transmission channels and affect different spread components.
For example, purchases of bonds issued by euro area non-financial corporations (NFCs) would probably have some direct pass-through effect to the financing costs of those firms which have the standing and ability to tap credit markets directly. And they could indirectly - through a sort of "trickling down" effect - make credit more abundant and less expensive for smaller firms by increasing the scope for banks to redirect credit provision from larger to smaller enterprises. Also, very low market-based funding costs might incentivise medium-sized firms to replace more expensive bank lending by issuing bonds.
But compared to other asset classes, the market for NFC bonds is relatively thin and quite heterogeneous across the euro area. This could imply that potential portfolio rebalancing and signalling effects would be contained. Furthermore, the large issuers which are active in this market are in all likelihood on average not financially constrained, which may limit the propagation of easier funding conditions to their decisions to expand capacity.
The transmission would work differently if we were to decide to buy bonds issued by euro area sovereigns.
This would be the only market where size would generally not be an issue. Interventions in this market would therefore likely entail a stronger signal that the ECB - consistent with our forward guidance - is committed to maintain an accommodative stance for an extended period of time. This signal should not only compress the risk-free curve further, it should also spill-over to inflation expectations, thereby helping to lower the real expected rate of interest.
We would also expect broad portfolio effects to materialise with sellers of government bonds willing to reduce their excess holdings of cash by bidding up the price - and hence lowering the yield - of alternative financial instruments other than those purchased by the central bank.
For banks, broader portfolio effects might entail a shift out of government bonds - or other fixed-income instruments - and into loan creation. This essentially works through banks' opportunity cost of lending. Our analysis shows that sovereign spreads in some Member States are still preventing lending rates from falling further. Remember that the existence of large spreads in the sovereign bond market was probably one of the main reasons why our 3-year LTROs had a limited effect in terms of stimulating credit supply by banks.
Although these spreads have narrowed sharply, there remain considerable differences in real - that is, inflation adjusted - yields across countries. Therefore, if purchases of government bonds were to reduce banks' opportunity cost of lending by lowering the return of other alternative investments, this might create the incentives for banks to extend credit to the private sector, in particular if banks regain confidence and start to reappraise the macroeconomic outlook as a result of central bank action. In addition, our presence as a buyer in the ABS and covered bond markets eases potential balance sheet capacity and liquidity concerns that go along with an increase in credit supply.
A precondition for these channels to unleash their full effects is a healthy banking system. The higher banks' capital ratios, the lower their bad loan exposure and the more transparent their balance sheets, the larger are the chances that the quantitative impulses of the central bank will get transmitted to the wider economy.
In this sense, the completion of the Comprehensive Assessment and the start of the SSM will provide important support to current and potential future efforts to revitalise sluggish lending in the euro area. In particular, increased clarity and transparency about banks' balance sheets, together with a better capitalised banking sector, will create a more supportive lending environment.
Of course, effectiveness is not the only criterion to assess the potential for new measures. In the euro area, the deliberation also needs to build on and factor in the institutional specificities of the euro area, including the boundaries set by the Treaty, notably with respect to the monetary financing prohibition and the no bail-out clause.
That is, we cannot simply import the experience from other central banks. We have to carefully design measures that are tailored to the institutional set-up of the euro area. Unlike other central banks, we are not dealing with a single fiscal authority but with a multitude of national fiscal authorities.
Let me conclude.
Since June the ECB has responded forcefully to counter growing risks of a too prolonged period of low inflation.
Taken together, our measures will have a sizeable impact on our balance sheet, which is intended to move towards the dimensions it had at the beginning of 2012. In the coming months, our measures will further ease the monetary policy stance more broadly, support our forward guidance on the key ECB interest rates and reinforce the fact that there are significant and increasing differences in the monetary policy cycle between major advanced economies.
But the latest euro area macroeconomic projections indicate lower inflation, accompanied by weaker real GDP growth and subdued monetary dynamics. In this context, early next year the Governing Council will reassess the monetary stimulus achieved, the expansion of the balance sheet and the outlook for price developments. We will also evaluate the broader impact of recent oil price developments on medium-term inflation trends in the euro area.
Should it become necessary to further address risks of too prolonged a period of low inflation, the Governing Council remains unanimous in its commitment to using additional unconventional instruments within its mandate. This would imply altering early next year the size, pace and composition of our measures.