The economic outlook and the ECB’s monetary policy: some key issues
Speech by Lucas Papademos, Vice-President of the European Central Bank, delivered at the Nomura annual Euro Conference “A challenging future for Europe”, Tokyo, 11 November 2004
The annual Euro Conference in Tokyo, which is now well established, provides a useful forum for examining various policy issues relating to the European economy, also from a global perspective. It is a great pleasure to have the opportunity to address you today and I thank you for the invitation. In my presentation, I would like to focus on two topics:
First, the state of, and outlook for, the euro area economy and the current stance of the ECB’s monetary policy. Addressing this topic requires me to assess, in particular, the implications of increased oil prices for the euro area economy and the single monetary policy.
Second, I thought it might be of particular interest to you if I address also the challenges posed for monetary policy by large swings in asset prices, especially housing prices. Given the Japanese experience in the 1990s, it is particularly appropriate to re-examine here in Tokyo the impact that large asset price swings can have on macroeconomic stability and the difficult choices for policy-makers confronting strong asset price dynamics. As you all know, these issues are not only of interest to Japan; they are of increasing relevance worldwide.
2. The economic outlook and monetary policy in the euro area
I will start by discussing the current state of the euro area economy and the medium-term economic outlook. The performance of the euro area economy this year is characterised by the ongoing recovery of economic activity at a pace somewhat faster than previously envisaged. Inflation, however, has remained above (though relatively close to) 2%. This outcome is not in line with our notion of price stability and is higher than the average inflation expected a year ago for 2004. The sharp rise in oil prices since the beginning of this year, by about 60% (in US dollar terms), has been the main reason for inflation being higher than expected earlier.
Let me elaborate a bit on the pace of the economic recovery, the factors that have contributed so far to the pick-up in growth, and the conditions and basic factors that are expected to influence the medium-term economic prospects.
Since the start of the economic recovery in the second half of 2003, we have seen robust growth rates in the first half of 2004, with real GDP increasing by 0.7% and 0.5% in the first and second quarters, respectively. The recovery in growth has so far been driven partly by buoyant external demand and partly by domestic demand. Private consumption increased moderately but investment was subdued.
Although short-term indicators have become more mixed over the past couple of months, the underlying determinants of economic activity remain favourable, and growth is expected to continue in 2005. On the external side, euro area exports should continue to benefit from positive global demand conditions in 2005. I admit, we have seen some moderation in global growth rates following a period in which the world economy has experienced its strongest dynamism in many years. We view this moderation, however, as an orderly return to sustainable, and still robust, rates of growth. On the domestic side, investment has thus far been subdued, partly due to weak construction investment. Looking ahead, investment should be supported by the global environment and the continuing very favourable financing conditions in the euro area. In addition, improved earnings and greater corporate efficiency gained through business restructuring should support investment. As regards consumption, we see scope for it to strengthen further, particularly once employment prospects have improved more visibly.
As with any assessment of the future, this outlook is surrounded by some uncertainty. In particular, there are risks to growth prospects stemming from unfavourable developments in oil markets. In this respect, let me emphasise a number of characteristics of the recent rise in oil prices. First, it is evident that recent oil price developments differ somewhat from previous oil price shocks, as they have been partly driven by buoyant global demand, and not only by constraints on the supply side. Second, oil prices, in real terms, have remained at a much lower level than in previous episodes. Third, the oil-intensity of production is lower in the euro area than it was in the past. And finally, in percentage terms, the oil price increase has not been as high as in previous oil price shocks. Despite this, however, oil price increases constitute a significant adverse shock to the euro area economy – through their direct impact on energy prices, their implications for global economic growth, and their adverse effect on the euro area’s terms of trade. If oil prices were to remain close to USD 50 per barrel, or even increase further, they would dampen the strength of the recovery.
With regard to inflation, oil price developments have had a powerful direct impact on the year-on-year change in the Harmonised Index of Consumer Prices (HICP), particularly in recent weeks. In October, annual inflation increased to 2.5%, according to a provisional estimate, from 2.1% in September. The oil price shock may feed through the economy and generate further indirect effects in the coming months. This is indicated by developments in producer prices, which showed a year-on-year increase of around 5% in September.
On the whole, however, the information available so far does not suggest that stronger underlying inflationary pressures are building up in the euro area. Wages have grown at rates of around 2–2.5% over recent quarters. This trend should persist for some time in view of the high level of unemployment in the euro area. Moreover, unit labour cost growth has been low (2.0 % in 2003, coming down to around 0.7% in the first half of 2004) and it can be expected that it will be contained in the coming quarters by moderate wage developments and labour productivity gains.
Notwithstanding these encouraging developments, there are still several upside risks to the outlook for price stability. These risks are mainly associated with oil price developments, possible new increases in indirect taxes and administered prices, and potential second-round effects stemming from wage and price-setting behaviour. The rise in measures of long-term inflation expectations in the euro area in the course of this year is also a cause for some concern.
In addition, our monetary analysis points to upside risks to price stability. Since mid-summer, the dynamics of M3 have strengthened again, mainly reflecting continued strong demand for transaction balances, which are included in the narrow aggregate M1, the most liquid component of M3. The acceleration of money growth is driven by the stimulative effect of the historically low level of interest rates in the euro area.
The low level of interest rates is also fuelling private sector demand for credit. In particular, the growth rate of loans for house purchase continues to rise in the euro area and is now approaching double digits. This goes hand in hand with relatively fast rises in real estate prices in many euro area countries. In recent months, the annual rate of growth of loans to non-financial corporations has also picked up.
Over the past few years, M3 growth has been very strong in the euro area, partly as a consequence of flight-to-safety portfolio shifts out of stock markets in the period of heightened stock market volatility between 2001 and 2003. As a consequence, there is now substantially more liquidity in the euro area than is needed to finance non-inflationary growth. This excess liquidity could pose inflationary risks in the future. Moreover, the persistently high excess liquidity and strong credit growth could become a source of unsustainable asset price increases, particularly in property markets. I will elaborate on this point later on in my presentation.
In summary, our assessment of the outlook for price stability, on the basis of the economic analysis and cross-checking with the monetary analysis, leads to the following conclusions: underlying inflationary pressures are still contained; there are still, however, a number of medium-term upside risks to price stability that must be monitored carefully; and for this reason, strong vigilance is required with regard to the materialisation of such risks. Against this background, the ECB’s Governing Council decided at its meeting on 4 November 2004 to leave the key ECB interest rates unchanged. On that occasion, we again noted that the level of interest rates in the euro area remains very low by historical standards. Indeed, at 2%, money market rates are the lowest since the end of the Second World War, and at 4%, long-term government bond yields are also well below long-run averages. The low level of interest rates, both in nominal and real terms, and the generally favourable financing conditions have been lending support to the ongoing economic recovery in the euro area.
Fiscal imbalances in the euro area have not improved in 2004, while the current account balance remained virtually unchanged. According to the European Commission, the average general government deficit in the euro area increased to an estimated 2.9% of GDP in 2004 (from 2.7% of GDP in 2003), while the current account is expected to record a surplus of close to 1% of GDP, slightly higher than in 2003. Although the Commission is projecting a decline in the total euro area fiscal deficit to 2.5% of GDP in the years 2005 and 2006, the slow pace of fiscal consolidation is a cause for concern, especially in view of the expected growth in the coming year and given the underlying imbalances associated with population ageing. Greater progress towards healthier public finances will strengthen confidence and support, rather than hinder, the ongoing recovery.
3. Economic effects and policy implications of higher oil prices
In this presentation of our assessment of the economic outlook and of our decision on the monetary policy stance, one aspect stands out which is of particular interest to investors, consumers and the financial markets at large – the recent sizeable rise in oil prices. This is an important issue, and one on which clear and convincing communication from the central bank is warranted.
Oil prices, which reached historical highs in nominal US dollar terms in the last week of October, are a key factor determining near-term prospects for economic activity and inflation, both globally and in the euro area. As I noted earlier, persistently high and rising oil prices have had a visible direct impact on consumer prices in the euro area, and inflation is likely to remain above the upper limit of our definition of price stability of 2% in the coming months. This is a worrisome development, although there is little indication that medium-term inflationary pressures are building up in the euro area.
What would be the potential impact of a sustained rise in oil prices on key macroeconomic variables in quantitative terms? For purely illustrative purposes, and using a variety of macroeconomic models, ECB economists have estimated the potential impact of a hypothetical permanent oil price increase of 50% on both inflation and output. According to these calculations, such an increase would add 0.3 to 0.6 percentage point to overall inflation in the euro area in the first year and 0.1 to 0.4 percentage point in the second year, before petering out in the third year (0.0 to 0.1 percentage point). The estimated impact on output would be, of course, negative, dampening real GDP growth by 0.1 to 0.8 percentage point in the first year; in subsequent years, the average cumulative impact suggested by these models would be roughly neutral (-0.3 to 0.2 percentage point in the second year; -0.1 to 0.4 percentage point in the third year).
Needless to say, these estimates are surrounded by a high degree of uncertainty related to the model specifications and other factors, for instance, the assumed initial cyclical position of the economy, the nature of the oil price shock, and the possible existence of non-linearities in the transmission of the oil price shock to the economy. More importantly, these (and other comparable) simulations are based on the assumption that the monetary policy stance is given; that is, the central bank’s interest rates are assumed to be kept constant and not to be adjusted in response to the impact of the oil price shock on inflation. Of course, this assumption is made for the purpose of isolating the “pure” impact of the oil price shock. In reality, monetary policy has an important role to play in influencing the effects of oil price shocks on the economy.
Oil price shocks present policy-makers with difficult choices as they simultaneously pose upside risks to inflation and downside risks to growth. In responding to oil price shocks, it is essential to understand that policy-makers cannot offset the real effects of oil price increases. An oil price increase triggers a loss in the economy’s terms-of-trade and implies a transfer of wealth from oil-importing countries to oil exporters. This change in the terms-of-trade (in relative prices) requires an adjustment of the real economy and must be absorbed by markets. The policy-makers’ aim should be to facilitate this adjustment by minimising inflationary pressures and aggregate output losses.
In reacting to oil price shocks, it is, therefore, important that policy-makers do not repeat the mistakes of the past by attempting to stabilise output growth at unrealistic levels. In the 1970s, such attempts led to a prolonged period of high and costly inflation, rising unemployment and sizeable fiscal deficits. Monetary policy should aim to ensure that inflation expectations are not adversely affected by the unavoidable “first-round” direct and indirect effects of an oil price shock on the price level and that they remain anchored to price stability. By preventing oil price shocks from having “second-round” effects on inflation expectations and on wage and price-setting behaviour, monetary policy can contain the unfavourable consequences of these shocks on both inflation and growth. It can prevent the fuelling of an inflationary spiral and the persistence of inflationary pressures, it can facilitate the real economy’s adjustment to the change in relative prices implied by the oil price shock, and it can thus help to minimise the associated output losses. This is why we at the ECB stress the need to be extremely vigilant against the materialisation of second-round effects that may result from a rise in oil prices. The ECB’s commitment to maintaining price stability in the euro area over the medium term implies that it will react promptly if there are indications that second-round effects are likely to emerge.
4. Asset prices and monetary policy
As has become clear from the previous discussion, monetary policy-makers may face very difficult choices in the event of large swings in important economic variables – like the price of oil and other raw materials – that can have a significant impact on the state of the economy. The same is true for large movements in asset prices, especially when markets – notably equity and housing markets – are seen to be moving out of line with what could be considered fundamentals. Such asset price misalignments, or “bubbles”, and their eventual correction, can potentially have severe economic costs.
Asset prices are important for monetary policy, not just because abrupt changes in asset prices may have severe implications for macroeconomic stability, but also because asset prices play a key role in the transmission of monetary policy. Through balance sheet effects and through their influence on the cost of capital, wealth, and confidence, asset prices affect spending decisions by households and companies. In addition, asset prices provide useful information to monetary policy-makers, notably about private sector expectations.
Over recent years, central banks and economic research have paid increasing attention to these issues. Indeed, the experience in Japan over the past two decades underscores how harmful bursts in asset price bubbles can be for macroeconomic stability, and how difficult it is for policy-makers to mitigate these costs. Indeed, there are quite a number of historical episodes when a collapse in asset prices triggered sustained deflationary trends and problems of financial instability.
When discussing asset prices and monetary policy, four questions come to mind:
First, is it possible to identify the emergence of an asset price bubble in the markets? And if so, how?
Second, assuming that a bubble has been identified, what can policy-makers do about it? What are their choices and dilemmas?
Third, what is the concrete situation in equity and housing markets in the euro area at present?
Finally, what role do asset prices play in the ECB’s monetary policy and what challenges could they pose?
4.1 How to identify a bubble?
Answering the first question is both easy and tricky at the same time: I think, we can all agree that it is very difficult to identify bubbles, not least since we need to distinguish asset price changes that are justified by changes in fundamentals, such as technological advances or changes in risk preferences, from those that are due to misalignments or “irrational exuberance” – to quote a prominent central bank colleague. For these reasons, central banks need to be – and they are – very cautious in identifying and pointing to the existence of a bubble. However, a number of indicators may help us to form an assessment about whether an asset price misalignment or a bubble is under way. A widely used indicator to gauge major stock price misalignments – an indicator which I am sure many of you employ in your professional activities – is to compare price-earnings ratios with historical averages. In the housing market, one method is to look at trends in the price-rent ratio relative to the historical averages, or at affordability (price-income) measures. All these indicators can be informative in judging the sustainability of asset price valuations.
In addition, a number of researchers have stressed that important complementary information for assessing the sustainability of asset price movements can be derived from monetary and credit developments. In particular, it has been shown that periods of high money and credit growth often give rise to boom-bust cycles in stock and property prices, which then causes high costs in terms of macroeconomic stability. Overall, we have several tools which could help us to detect bubble phenomena. While individual tools are subject to caveats, applying a broad range of methodologies should help limit the risk of making mistakes in the assessment of asset prices.
4.2 What can central banks do about bubbles?
Coming to my second question, how should central banks react if they see the risk of an emerging asset price bubble? The options range from not reacting at all to consciously trying to prick an emerging bubble. Obviously, attempts to deflate an asset price bubble are likely to require substantial interest rate increases which, in turn, could adversely affect economic activity and inflation. In view of the difficulties in identifying a bubble, it seems established wisdom today that central banks should be exceedingly careful in dealing with asset bubbles.
There is, however, a more subtle way: to cautiously consider “leaning against the wind” when an asset price bubble is developing. Indeed, some academics have argued that, by accepting some short-term deviation from the pursuit of its stated objective of price stability, a central bank may improve the conditions for maintaining price stability over the medium to long-term. If a potentially costly bubble is firmly under way, then a somewhat tighter monetary policy stance than would otherwise be needed could be warranted.
Certainly, there are also risks involved in “leaning against the wind”. First, as I have already mentioned, there can never be full certainty as to whether an asset price bubble is under way. Second, by influencing market psychology, a monetary policy tightening could trigger dynamics that are difficult to foresee and may be hard to control ex post.
Effective communication, as always, is crucial, especially in situations of heightened uncertainty. The central bank can consider adjusting its communication pre-emptively in order to discourage the formation of overly optimistic expectations regarding future returns and/or in order to signal concerns about too low risk premia. In any case, monetary authorities, through their communication, should avoid appearing complacent, thus possibly contributing themselves to unrealistic expectations and asset-price dynamics that are not in line with fundamentals.
Finally, central banks should not be seen as reacting asymmetrically to asset price movements; that is, as only reacting to mitigate the consequences of a fall in asset prices. Obviously, such asymmetric behaviour can lead to moral hazard in the market, with the central bank becoming a potential contributor to excessive asset price fluctuations, if its actions induce private agents to take higher risks than would otherwise be the case.
4.3 What is the situation in the euro area housing and equity markets?
After these rather general considerations, let us look at the current situation in the euro area. What are the underlying factors fuelling recent developments in asset prices? Are these developments a cause for concern and what are the associated risks? Let me first focus on residential property markets. As already noted, growth rates in euro area property prices have been rather dynamic in recent years. After relatively modest price increases of less than 3% between 1993 and 1998, residential property prices on average for the euro area as a whole have been rising at annual rates of around 6-7% in each of the past four years. In real terms, this corresponds to growth rates of 4-5%, which is relatively high by historical standards. Differences across euro area countries have been considerable: while in Spain, France, Ireland and Italy, residential property prices were rising on an annual basis by between 11% and 17% in 2002 and 2003, they have been falling slightly in Germany and Austria in the last four years.
The dynamism of property prices in several euro area countries is very much related to the very low levels of interest rates brought about by the introduction of the euro. For several euro area countries, the adoption of the euro marked a significant fall in real interest rates, as it implied a “regime change” towards a more stable macroeconomic environment. The fall in interest rates, in turn, allowed households to borrow more and accumulate more debt. As a consequence, household indebtedness in the euro area has increased significantly in recent years. In 2004 it is estimated at 54% of GDP, up from 46.5% in 1998 and 43% in 1995.
The rise in household debt has not, however, been accompanied by higher interest payments as a ratio of disposable income in the euro area as a whole. This ratio amounted to 4.6% in 2003, similar to the 4.8% observed in 1998. However, the strong rise in house prices and the substantial increase in household indebtedness have raised some concern. Given the differentiated path of house price increases across countries, this concern is obviously more relevant and stronger in some countries than in others.
The concern relates in particular to the likelihood that some households may be underestimating the risks they face. These risks are associated, first, with the higher sensitivity of households to changes in interest rates. In this regard, we have seen a tendency in some euro area countries to increasingly focus on borrowing on conditions which are linked to short-term interest rates. For many households this seems to be an attractive option as it implies a rather low interest rate burden notably at the beginning of the contract. What is often overlooked, however, is the fact that the currently low interest rates could well change over time.
A second risk relates to future house price changes. Some valuation measures for property prices, such as the price-to-rent ratio, are now 10-15% higher than their average levels during the period 1995-2003 in the euro area as a whole. These valuation measures have even reached levels as high as 17%. Against this background, concerns have been expressed that substantial house price increases in some countries entail risks of a reversal of this trend that may currently be underestimated.
It is not the first time we have been confronted with pronounced movements in asset prices in the euro area. Indeed, current movements in house prices are small when compared to some swings in stock market prices over the last six years. We have witnessed rather substantial share price fluctuations resulting from the new economy euphoria and the subsequent disillusionment. And we have, of course, tried to assess the impact of the dramatic stock market decline from the peaks in 2000 on macroeconomic conditions and financial stability.
In this regard, it mattered that, partly linked to the introduction of the euro, the outstanding amount of shares substantially increased in the euro area. For non-financial corporations in the euro area, the annual net issuance of shares increased from 1.7% of GDP in 1995 to 7.3% of GDP in 2000. At the same time, during the late 1990s, the distribution of shares across the euro area population broadened significantly, implying that more households invested in shares in the late 1990s than before. Corrected for valuation effects, the amount of households’ holdings of shares and mutual fund shares increased between 1995 and 2003 by 78%.
Against this background, stock prices have become more relevant to economic conditions in the euro area in recent years. And indeed, when stock prices started to fall, we could see that this had a notable dampening effect on the euro area economy in the period 2001-2003.
4.4 What role do asset prices play in the ECB’s monetary policy?
After this description of the current situation, the crucial questions are: how do asset price developments enter into the assessment of the outlook for price stability and do how they influence the policy-formulation of the ECB? Recently, much attention has been devoted to the question of which monetary policy strategies are optimal in an environment of strong asset price fluctuations. With regard to inflation targeting strategies, it has been acknowledged that if financial imbalances accumulate and there is, for example, substantial uncertainty about the sustainability of asset price movements, it is not advisable to strictly pursue an inflation forecast for consumer prices over a horizon of one to two years. In such circumstances, it may be better instead to set interest rates with a view to a time-frame extending well beyond conventional forecast horizons. The ECB’s monetary policy strategy has some features that can become quite useful in this respect:
First, from the beginning, the ECB has adopted a “medium-term orientation”. This implies that policy is not set for a time horizon of one or two years. Rather, the ECB pays due attention to the need to take into account the entire horizon over which monetary policy and macroeconomic shocks impact on the state of the economy. This lengthening of the policy horizon helps to trace out the likely macroeconomic impact of a financial misalignment.
Second, the monetary policy strategy of the ECB is broad-based, implying that a whole range of several indicators and models are employed and cross-checked, thus making the assessment and policy response to financial imbalances robust. In particular, the commitment to monitor and assess money and credit developments – besides being instrumental in identifying longer-term risks to price stability in general – may help to guard against ignoring asset price bubbles. Indeed, growth rates of money and credit which are persistently in excess of those needed to sustain non-inflationary economic growth may, under certain circumstances, provide early information on emerging, and potentially destabilising, asset price bubbles. In this respect, money and credit growth acts as a permanent admonition for vigilance in the face of unusual movements in asset prices.
From what I have said, it has become clear that it is important to monitor asset prices and to try to avoid the emergence of unsustainable, pronounced movements in asset prices which may subsequently lead to sharp recessions. Obviously, there cannot be a mechanical link between asset prices and monetary policy instruments. Central banks, however, should take asset price movements into account, especially if large swings in asset prices imply long-term risks to financial and economic stability.
5. Concluding remarks
As you can see, there are a number of key concepts which I have emphasised in my speech today: “strong vigilance”; “careful monitoring and analysis”; “medium-term orientation”; “prompt reaction”. These are central elements of our approach to monetary policy-making, both at the current juncture and in connection with the specific topic of asset price developments that I discussed in the second part of my address. We live in a world of considerable uncertainty, not just in terms of economic shocks – which by definition are impossible to predict – but also in terms of our knowledge of how the economy functions, how economic agents form expectations and react to events and changing conditions. Under such circumstances, the central bank’s unambiguous commitment to attaining its objectives is an indispensable anchor. A credible central bank with a clear commitment to price stability can, and must, provide such a haven of trust and confidence within the complex and uncertain web of interrelations in a modern economy. This is precisely what the ECB is seeking to do. Our track record over the past, almost six, years shows that we have been rather successful.
Thank you very much for your attention.