Peter Praet: Monetary policy transmission in the euro area
Policy address by Mr Peter Praet, Member of the Executive Board of the European Central Bank, at the SUERF Conference "Global Implications of Europe s Redesign", New York City, 6 October 2016.
I would like to thank Claus Brand for his contributions to this speech.
All major central banks have faced an extremely challenging economic environment since the onset of the global financial crisis in 2008. They have had to contend with the steepest collapse in output since the 1930s; with the aftermath of a combined banking and debt crisis, implying a prolonged period of deleveraging and an atypically slow recovery of output towards potential; and, more recently, with a series of global disinflationary shocks emanating from persistent capacity underutilisation in advanced economies, the slowdown in emerging market economies and collapsing energy and commodity prices, threatening to unmoor inflation expectations. And central banks have had to confront these challenges while operating for many years near the effective lower bound on interest rates.
These constraints have required central banks to deploy unconventional measures with unprecedented intensity to stabilise inflation expectations and secure medium-term price stability. But in the euro area, monetary policy has also faced another type of challenge, linked not so much to the stance of monetary policy as its transmission. This was a challenge born in particular of weaknesses in the euro area's institutional framework. Lacking a common framework for bank recovery and resolution, banking sector repair in the euro area proceeded in a slower and more piecemeal fashion, notably in comparison with other large economies like the US. And without institutions to efficiently manage a sovereign debt crisis, the euro area experienced a severe widening of sovereign spreads in 20102 and a second recession, both of which fed back into further strains on bank balance sheets. Given the central role played by banks in the transmission process in the euro area, these factors fundamentally impaired monetary transmission for long periods of the crisis.
This perspective explains why, for most of the period surrounding the sovereign debt crisis, our policy measures were primarily focused on addressing frictions in the transmission process. A major burden was placed on the central bank to compensate for an incomplete monetary union, yet the ECB was still able to provide significant macroeconomic stabilisation in spite of these adverse conditions. But that episode inevitably set back the recovery: banks in large parts of the euro area became less willing and capable of keeping credit flowing to the real economy, producing a vicious circle of contracting credit growth and weak demand dynamics. While the US had already recouped its pre-crisis real GDP levels in 2011, the euro area continued to face a yawning output shortfall. Accordingly, headline inflation began to drift downwards, owing both to global energy and commodity price developments and to ongoing weakness in the core components.
Hence, while US monetary policy began to gradually normalise, in June 2014 monetary policy in the euro area embarked on a renewed phase of expansion, aimed both at enhancing monetary policy transmission, in a context of continued bank deleveraging, and reinforcing policy accommodation in view of persistently weak inflation. With room for manoeuvre on the main refinancing rate limited, the ECB has articulated this stance in a series of new unconventional measures. What I would like to discuss in my remarks today is how these measures are transmitting through the financial system and economy to secure our price stability objective - their main channels of financial transmission, how those channels are affected by structural shifts in financial intermediation, and how they feed through to output and inflation in the real economy.
Transmission channels of unconventional policies
The ECB is delivering its accommodative policy stance through a set of complementary instruments. These consist of our targeted longer-term lending operations (TLTROs), our asset purchase programme of public and private securities (APP), and our policy of charging zero interest rate on main refinancing operations and a negative rate on excess reserves. The policy package is also underpinned by our outcome- and date-based forward guidance: that the duration of asset purchases will be conditional on a sustained adjustment in the path of inflation, and that interest rates will stay low for an extended period of time, stretching well beyond the horizon of the net purchases. In combination, we see these measures as being transmitted via three main channels: direct pass-through, portfolio rebalancing and signalling, with these channels reinforcing each other.
First, some of our tools are specifically designed to act on those financial instruments that have an immediate influence on the setting of the price of credit by financial intermediaries - the direct pass-through channel. I am referring specifically to the TLTROs and our purchases of asset-backed securities (ABS) and covered bonds under the APP, a group of measures we have referred to as our "credit easing" tools. These tools act on specific portions of banks' liability structures - central bank credit and wholesale funding, respectively - where the connection with the pricing of bank credit is the closest.
ABS and covered bonds are packages of bank loans. So, by exerting downward pressure on the market interest rate paid by originators of ABS and covered bonds, our measures aim to foster an economic arbitrage: banks are encouraged to create more loans with a view to re-packaging them and selling them on, thereby "cashing out" the spread between the interest rate at which the securities are sold to the ECB and the interest rate charged on the underlying credit. The TLTROs work similarly by design. Under this targeted term refinancing operation, banks - all banks - can borrow at the interest rate on the Eurosystem's deposit facility, but only on condition that the borrower can demonstrate strong performance in loan origination. Here, the intention is to introduce more competition in the market for bank loans. More competition squeezes unit lending margins and the level of borrowing costs for the real economy.
This "credit easing" mechanism is expected to compress the spread between the financing conditions set in the capital market and the borrowing conditions faced by the individual borrowers in the bespoke market for individual loans. But what is the mechanism that keeps financing conditions in the open market well anchored at sufficiently accommodative levels in the first place? That mechanism is set in motion by our interventions in the sovereign bond space under the APP. It is by exerting downward pressure on the sovereign curves that we ensure that the bedrock pricing kernels used in each country to price the whole spectrum of local assets and credit, namely the term structure of sovereign interest rates, is sufficiently supportive for the economic recovery to firm, and for inflation to revive durably.
Such interventions in the public securities segment rely on the second transmission channel - portfolio rebalancing - as a key avenue of propagation: the compression of returns in that market induces investors to move up the risk and maturity ladder, bidding up assets with higher risk-adjusted returns. Banks are again a key player in this chain of transmission. Our purchases of sovereign bonds depress the term premium and simultaneously trigger a rebalancing of bank balance sheets towards an expansion of asset holdings and lending, so as to offload the cash reserves that are created in the process. This incentive to move out of cash reserves is intensified by our negative rate policy.
As for the third channel - signalling - both our purchases and our rate forward guidance are effective instruments in influencing expectations. The credibility of promises to follow a certain course of action for setting the policy rates in the future are almost certainly enhanced by provision of asset purchases today, as these purchases are a concrete demonstration of a desire to provide additional stimulus. Conversely, the net stimulus provided by asset purchases depends in part on expectations of how the central bank will adjust short-term interest rates in the future in response to the stronger real activity and inflation sparked by lower term premia in the near term.
Structural issues in monetary policy transmission
These various transmission channels, however, are not static; they depend also on the structure of financial intermediation and how it evolves over time. Given that credit intermediation has been largely bank-based in the euro area, enlisting banks as the main propagators of stimulus, as we did when we launched a series of LTROs between 2008 and 2012 and, later on, when we decided to initiate the TLTROs, was a natural choice for us. At the same time, one should not ignore the new reality in which we see non-banks growing in importance for financial intermediation, also in the euro area.
The structural shift in finance away from bank lending to credit intermediation through financial markets has been ushered in by financial innovation, but it is also a consequence of the global financial crisis and the changes in financial regulation that have followed it. The euro area is no exception to this development: while remaining far below corresponding levels in the US, the share of non-banks, such as insurance companies, pension funds and asset management companies in financing non-financial corporations in the euro area has risen, though from a very low level. This development reflects longer-term structural factors, such as regulatory arbitrage, but also the low level of interest rates driving search for yield and the weak recovery of bank lending in the wake of the crisis.
How would monetary transmission in the euro area change if we moved all the way to a capital market-based system? That banks may become relatively less important in financing households and firms would not necessarily mean that monetary policy transmission would be weakened. To start with, there are inherent limits in firms' ability to substitute bank financing for market-based financing, and the scope for such substitution is likely to be smaller in the euro area - with a comparatively stronger presence of small and medium-sized enterprises (SMEs) in its industrial fabric - than elsewhere. But even abstracting from such market imperfections, one could presume that bank lending standards and conditions in securities markets may reinforce each other, supporting the effectiveness of monetary policy transmission.
As non-banks operate under less stringent regulatory constraints, they might step into banks' lending activities where bank balance sheets are already constrained. In doing so, they can even mitigate the effects of financial crises on the provision of credit to the broader economy. Likewise, their asset portfolio and risk appetite may be more sensitive to interest rate changes, making them more responsive to changes in monetary policy - although, admittedly, this enhanced sensitivity to monetary policy is something we would ascribe more to securities brokers and dealers, whose leverage tends to be more pro-cyclical, rather than to insurance companies or pension funds, which act more as long-term investors.
Certainly, there would be a trade-off. On the one hand, a higher share of arm's-length finance in total intermediation could well increase the share of borrowers in the real economy that benefit from very low, or even negative, short-term interest rates, compared with a system that is based by and large on banks. This is because non-bank lenders - compared to banks which finance their loans predominantly with sticky interest rate retail deposits - can more easily fund a larger share of their exposures in the capital market, where the short-term interest rates faced by prime institutions are negative, and thus have more scope for passing through those negative interest rates to their clients. But on the other hand, this would come with some financial stability risks: a larger share of wholesale market finance in conditions of low interest rates invites more maturity transformation, which may expose the financial system as a whole to sudden stop risks.
One may also believe that portfolio rebalancing and risk-taking may play a comparatively more powerful role in transmitting the type of large-scale asset purchases that the Federal Reserve and other central banks, including the ECB, have conducted. For example, in the case of the US, due to the more intimate connection between capital market conditions and the financing terms faced by US borrowers, successive rounds of Fed large-scale asset purchases could probably count on a stronger reinforcing role through exchange rate effects and the impact on global yields. This has worked in both directions: in the context of the "taper tantrum" in 2013, a large number of advanced and emerging economies experienced the potent global effects of the portfolio rebalancing channel being partly reversed, with medium-term money market interest rates moving up in lockstep with US yields on a global scale.
But overall, there is a range of common factors affecting banks and non-banks in similar ways in monetary policy transmission: the valuation of a broad range of financial assets, risk appetite and macroeconomic developments. A decrease in risk premia, for example, through its impact on funding conditions and portfolio valuation, is likely to support balance sheet expansions in both sectors, easing the borrowing conditions of non-financial corporations. Banks here operate in the same fashion and on the basis of the same relative-value logic as non-bank market participants: a policy-induced decline in fixed-income securities earnings encourages banks into the same type of portfolio re-allocations that affect investors in the capital market. The speed and amplitude of reallocation will probably be different, though. Non-banks may react to changes in market prices more rapidly than banks and banks may well entail a higher degree of inertia in investment patterns, with lending conditions becoming disconnected from the terms on which they borrow at some critical junctures.
At any rate, the interplay between different financial intermediaries and transmission channels in complementing traditional bank lending needs to be analysed and understood in an integrated way. Assessing this structural shift has been a matter of quantification as much as of judgement.
Impact of easing measures on financing conditions and macroeconomic developments
What is already clear, however, is that our package of unconventional measures has had a substantial impact on euro area financial conditions - an impact which was indeed observable soon after their launch in June 2014. Bank lending rates that had remained stubbornly high in the preceding years started trending downwards already in the third quarter of 2014. Borrowing costs faced by households and companies have since dropped to unprecedented lows. And this improvement is also evident if we look at the dispersion of bank-individual lending rates (for loans to non-financial corporations). The country distribution of interest rates charged by banks has returned to levels comparable with pre-crisis norms.
The generalised decline in bank lending rates has a twofold expansive effect. The first is that it frees up purchasing power that was previously locked up in high loan servicing costs. Whether it is because loan rates are floating and are reset according to a particular money market benchmark, or because borrowers are able to refinance outstanding credit at a much more attractive fixed interest rate - and possibly also to extract housing equity in the re-negotiation process - the end result is that borrowers receive a boost in disposable income that they can use to fund current consumption or expand investment.
The second effect that is associated with a decline in lending rates of the magnitude we have observed in recent years is more general-equilibrium in nature. Because the interest rates charged on loans to non-financial companies have been falling so sharply, more investment projects become profitable and the demand for loans expands. Note that bank unit margins may well decline in this process. But overall margins may even rise if the demand for loans is sufficiently responsive to the interest rate, as banks can collect lending margins on an expanded pool of assets. Indeed, the pace of contraction of loans growth at country level, which was massive at the start of our policies in summer 2014, started to decelerate in 2015, and loans to households and firms have returned to positive growth rates recently. This being said, the low sensitivity of real investment to borrowing conditions has declined everywhere in the aftermath of the crisis.
Our Bank Lending Survey (BLS) confirms that the stimulative effect of bank competition on loan demand is in fact working. As regards the role of TLTROs, banks participating in those lending operations have felt the pinch of stronger competitive pressures comparatively more than banks abstaining from TLTRO participation, a feature clearly borne out by the BLS as well. As regards the role of APP and the negative rate on excess reserves, banks having received comparatively more net inflows of excess liquidity report stronger reductions in the margins on loans to enterprises. This is testament to the fact that the negative remuneration of excess reserves discourages liquidity hoarding and reinforces the inducement for banks to rebalance their asset composition towards loans and securities with longer maturities, confirming our expectations that investors move up the maturity and risk ladder.
Overall, the BLS reported a large share of banks decreasing lending rates and easing credit standards in the wake of TLTRO, APP and the negative interest rate. Importantly, both our BLS and SAFE attest to improved credit conditions for SMEs, which form the backbone of the euro area's corporate sector. And the effectiveness of our measures is highly visible in bond yields across sectors, too. Sovereign yields have fallen and flattened out concurrently with the money market yield curve. Yields on bonds issued by banks and non-financial corporations have decreased by similar amounts. This broad-based asset repricing has also affected the stock market and the euro exchange rate.
Turning to the macroeconomic effects, if we look at the real-time flow of high-frequency macroeconomic news and at the evolution of long-term market-based inflation expectations, two observations stand out. The credit easing component of our package and, later on, the APP announcement appear to have contributed to a turnaround in business sentiment and halted a creeping downtrend in inflation expectations. It is hard to see how, without the marked improvement in financial conditions produced by our measures, growth in employment, investment and consumption, which was teetering on the edge of another slowdown in 2014, would have stabilised around the moderate, yet steady, path we see today. Moreover, the euro area economy would not have weathered the repercussions from a major slowdown in global trade, a potentially disruptive loss of confidence in the stability of the international economy and, more recently, the Brexit referendum with the resilience that it has demonstrated, if these measures had not been in place.
Indeed, we have to recognise that much of the progress registered in the past, and a large fraction of the progress towards a higher inflation that we project for the future, is due to our monetary policy stance. As such, the gradual return of our economy to a balanced growth path in conditions of price stability that we foresee in our forecasts would most likely be stalled and reversed if the monetary expansion that is presently embodied in credit conditions were to be withdrawn prematurely. Our measures are estimated to contribute to increasing real euro area GDP growth by more than one and a half percentage points cumulatively between 2015 and 2018. Likewise, the estimated contribution of the measures to the projected return of inflation to levels closer to our objective is significant, with an impact on the annual inflation rate, on average, over 2016 and 2018 of about half a percentage point.
Let me conclude.
A key question in designing the modalities of our unconventional measures, over the various stages of the crisis, has been how to overcome impairments in monetary policy transmission as a precondition for the ECB to be able to deliver on its mandate. In the wake of our measures, financing conditions have improved significantly. They have supported a recovery of credit provision and shielded the euro area economy from adverse global shocks.
As I have discussed elsewhere, we need to be attentive to the possibility of adverse side effects of the low interest rate environment1 and take this into account in the calibration of our monetary policy toolkit so as to preserve the very substantial amount of monetary support.
1 See speech by Peter Praet, "Monetary policy and the euro area banking system", at VII Financial Forum organised by Expansión and KPMG, Madrid, 4 October 2016.