The Growth of Pension Funds and Financial Markets: Implications for Central Banks

Speech by Lorenzo Bini Smaghi,
Member of the Executive Board of the ECB
Dinner speech IMCB conference on “Dealing with the New Giants”
Geneva, 4 May 2006

Ladies and Gentlemen,

It is a great pleasure for me to introduce this conference on pension funds, which aims at finalizing this year’s Geneva Report.

We all agree that pension funds are becoming key players in financial markets. Central banks should thus look at pension funds with great attention. This includes the European Central Bank, given the scope for growth of this segment in the euro area (1 trillion US dollars of outstanding assets in 2004, compared with over 11 trillion in the United States).

Tonight, I would like to touch on a few issues related to how the development of pension funds may impact on the conduct of monetary policy. I will first describe how institutional investors might be changing the functioning of financial markets. I will then consider the implications of these changes on the role of asset prices as information variables for monetary policy. Finally, I will touch upon the possible effects of these changes on the transmission of monetary policy.

Pension funds and financial markets

There can be little doubts that the growth of pension funds is bound to affect the structure and functioning of financial markets, in different ways.

In principle, a world in which financial market are dominated by sophisticated investors can bring important advantages. Efficiency, depth and liquidity can be greatly enhanced.

There may also be risks, and I will concentrate mainly on these.

A first risk is that asset prices get misaligned, due to portfolio managers’ behaviour. Institutional investors are prone to the so-called agency problem, i.e. the difficulty of aligning the incentives of savers with those of managers. For pension funds, in particular, the shortening of the time horizon associated with the progressive aging of society might create incentives for asset managers to be myopic.

There is ample economic literature on the fact that “agents” may be prone to moral hazard, and that there is no “first best” contract for delegated portfolio management that can fully eliminate this risk. In general, it cannot be assumed that savers’ preferences are fully reflected by delegated portfolio managers. In fact, contractual arrangements tend to assess managers’ performance with respect to some heuristic benchmark, defined over a relatively short time-horizon. If benchmarks play a key role in portfolio management, a natural tendency to herd may emerge among institutional investors.

As is well known, herding can be a recipe for asset misalignments, in particular over the short term, and for the emergence of self-fulfilling price setting.

A second risk is market concentration. Given the substantial economies of scale to be exploited in the industry, few key players - what the Report calls “the New Giants” - may emerge over time, dominating the market. It is striking to note, in this context, that already now 40 per cent of outstanding Government bonds are held by pension funds. Concentration may reduce the efficiency of the market and increase its fragility, in particular in the face of contagion.

The combination of the two risks mentioned above - myopic behaviour and market concentration - might further fuel the risk of herding behaviour. The fewer the portfolio managers, the easier it may be for them to “coordinate” their expectation and their market positions, independently of underlying economic fundamentals. To use Keynes’ analogy with the beauty contest, the fewer are the judges, the easier it is for them to reveal their preferences to each other before expressing their choice. Asset managers might thus become more and more focussed on trying to anticipate other managers’ expectations, including the central bank short-term behaviour, rather than looking at fundamentals.

What do the above risks entail for central banks?

Financial markets are important to central banks for two main reasons. First, they provide policy-makers with information on private sector expectations about fundamentals (growth and inflation). Second, they are an important channel of transmission of monetary policy. If the functioning of financial markets is influenced by the development of institutional investors, in particular pension funds, in a way in which the above mentioned risks emerge, central banks’ modus operandi is directly affected.

I will consider in turn the two functions of financial markets, as information provider and as channel of transmission.

Financial Markets as Information Provider for Central Banks

If the fixed income market is dominated by few institutional investors seeking to beat a benchmark, there is a risk that asset prices might reveal little about fundamentals. Asset prices might tend to be determined mainly on the basis of other investors’ expectations and on the anticipation of central bank policy over the short term.

This raises a dilemma for central banks.

On the one hand, if market prices tend to reflect more the expectation of future central bank action than underlying market fundamentals, there is little information that central banks can extract from these prices. If central banks give excessive importance to market prices, there is a risk of the "dog chasing its tail", to use Alan Blinder's words:[1] financial markets look at the central bank and the central bank looks at financial markets, and both lose sight of the underlying factors determining inflation.

On the other hand, central banks want to be highly predictable because this improves their credibility, the transmission of monetary signals and ultimately the functioning of markets. It is therefore highly desirable that market participants anticipate adequately central bank action. This is why market participants seek guidance from central banks for the formation of asset prices. If guidance is provided on a continuous basis, market participants’ incentive to make their own assessment on underlying fundamentals tends to fade. If guidance is not provided, there is a risk that financial markets’ expectations deviate from central banks’ intentions, with possible major ex-post price adjustments that may undermine central bank predictability and credibility.

There is clearly a problem of time consistency in the optimal communication, or guidance, that central banks can provide to markets. A trade-off needs to be made between central banks’ desire to closely guide market behaviour and the need to let markets make their own assessment and leave them the responsibility for their own expectations.

Over time, a consistent policy might be to guide markets not by revealing specific policy intentions month after month, but by providing information on the key aspects of policy decisions, i.e. the definition of the policy objective, the policy strategy, the analytical background and the way in which incoming data are assessed and projections are made. Ex post explanations of policy decisions are also essential to enable market participants to understand how past policy decisions have been taken so that they can understand how future decisions will be made.

Market participants will of course always try to tempt central banks by asking them for confirmation or denial of the prevailing market expectations. This temptation may be stronger the more market participants are prone to myopic behaviour and the greater is market concentration. However, central banks might want to resist this temptation and explicitly correct or confirm market expectations only under exceptional circumstances.

This approach entails that market expectations may occasionally deviate from the assessment of the central bank. Market participants may thus have to reassess their expectations, as it happened for instance in the euro area in March-April 2006. This can lead to some disappointment, possibly speculative losses and, more worrying for the central bank, a potential loss of reputation. What is essential in these circumstances is that the central bank sticks to its communication strategy and continues to explain in a transparent way the analytical framework underlying monetary policy decisions.

Financial Markets and the Transmission of Monetary Policy

A second question of interest for central banks is whether the emergence of large pension funds affects the transmission of monetary policy on output and inflation.

In a textbook model, monetary policy affects asset prices via the discount rate, which determines, together with the risk aversion of the representative investor, asset yields which affects households’ and firms’ behaviour. If, however, asset prices are set on the basis of short-term benchmarking, the relationship between the policy rate and households’ and firms’ behaviour may become non-linear and unpredictable.

The low interest rate situation which has characterised the recent past can further blur the relationship between policy rates and asset prices. In particular, the compression of risk premia in a variety of markets and the spectacular rise in asset prices in some segments, notably in the real estate market, might be examples of such behaviour. A hypothesis is that the transmission of monetary policy on asset prices might be stronger at low interest rate levels, due to the fact that benchmarks tend to be set on the basis of some historical average. The existence of such benchmarks might make managers more eager to beat them and hence readier to take on risks, which in turn amplifies the rise in asset prices and the compression of asset yields.

The emergence of pension funds may in fact itself contribute to an environment of low interest rates, especially at the long end of the curve, reflecting the increasing demand for fixed income assets. This may create phenomenon like “conundrum”, in which long term rates behave in a way not fully consistent with movements in short term rates. Such conundrum may impact directly on the transmission mechanism of monetary policy.

These are all areas where further research is needed.

Finally, let me also mention in passing that the rising importance of institutional investors and the associated changes in the financial structure has an impact on monetary analysis, which is currently largely based on information provided by banks. On the one hand, while institutional investors also hold money, their liquidity and asset allocation may significantly differ from those of households and firms whose funds they often manage. On the other hand, institutional investors also directly engage in lending to households and firms and their activity may thus substitute for bank lending. For instance, in the US, we have observed a rise in both sides of households’ balance sheets. Institutional investors, including pension funds, have strongly contributed to this development. This makes it more difficult to detect whether an increase in some parts of household liabilities, for instance bank credit, is motivated by consumption behaviour (with an impact on aggregate demand and potentially inflation) or by asset-liability management.

In the euro area, theses developments are still at their infancy. However, if a trend towards a US model emerges, we can expect substantial changes in the financial structure of the household sector and in the way financial instruments are used. Going forward, the information provided by monetary financial institutions might have to be complemented with timely balance sheet information on institutional investors, to properly take into account the triangular relationship between banks, institutional investors and households and the possible substitution between bank-based and institutional investors’ financial transactions. Such structural developments need to be properly considered when evaluating the information signals coming from credit and monetary developments for the assessment of the medium-term outlook for price stability.

To sum up, the rise of pension funds and more generally of institutional investors is an overall positive development for financial markets. It has nevertheless a series of implications, in particular for the conduct of monetary policy, which requires further analysis.

[1] A. Blinder, Central Banking in Theory and Practice, MIT Press, Cambridge, Mass, 1998.

Speaking engagements

Media contacts