Executive summary Observed trends in institutional investors' behaviour Institutional investors are becoming more important in global financial markets, with their assets under management rapidly catching up with those of the banking system. Institutional investors help to ensure deeper and better functioning markets, thus contributing to a more efficient allocation of savings, and their growth may help to counter the decline of household saving ratios associated with ageing populations. Some of this growth will be stimulated by public pension reform, including moves to funded occupational pension systems. The growth in turn is likely to be concentrated in the defined contribution (DC) pension sector because the majority of countries that are moving to funded pension systems are adopting a structure based on a DC arrangement. Global institutional investors have also substantially increased their exposure to emerging market economies (EMEs) in recent years. Domestic institutional investors in EMEs, although small in an absolute sense, are significant relative to the size of local markets and have considerable growth potential. Different objectives and strategies determine the investment behaviour of different institutional investors. The asset allocation of insurance companies or defined benefit (DB) pension schemes that bear investment risk could be expected to be different from that of mutual funds or DC pension schemes, in which the risk is directly borne by the individual. Asset-liability management strategies or a need to close funding gaps may be characteristic of the investment behaviour of the first group of investors, but not of the second. Alternative investments have become increasingly popular among traditional institutional investors such as life insurance companies and pension funds, but still represent only a small proportion of total assets. However, from the hedge fund viewpoint, these investments constitute a significant source of growth for the industry. Effect of regulatory and accounting changes Regulatory developments have long been a factor in the development of institutional investors and their asset allocation strategies. The Working Group was mandated to assess the effect on investment behaviour and financial markets of specific regulatory and accounting changes affecting pension funds and insurance companies. These regulatory changes were motivated, at least partially, by the 2000-02 equity downturn, which exposed some serious weaknesses in the regulatory frameworks affecting traditional institutional investors with long-term liabilities and offering guaranteed returns in many countries. These regulatory policy changes are of a global nature. However, their consequences are likely to differ between countries, partially as a result of differences in national conditions, as well as across different types of institutional investor. The main effect of reforms will be to provide incentives for DB pension funds and insurance companies to reduce their risk profile, in part by making risks more visible. This process can take many forms, but generally involves either the transfer of investment risk to households or the adoption of investment strategies that directly incorporate liabilities into asset allocation decisions. Underfunded DB pension funds and insurance companies selling guaranteed products will be encouraged to reduce their risk profiles. Overfunded DB pension funds or well capitalised insurance companies will be less constrained by regulatory reforms in the risks they take. However, even if a DB plan is currently overfunded, the sponsoring firm's concerns about the risk of becoming underfunded in the future may encourage greater asset-liability matching (ALM). The adoption of ALM techniques by institutional investors may involve shifts in asset portfolios from equities to long-term conventional and index-linked bonds, whose financial characteristics more closely resemble liabilities in terms of duration and the cash flow of obligations. Institutional investors may also seek to transfer risk to the household sector. In shedding risk from their balance sheets, they will be following a similar strategy to the banking sector. Of course, households are the ultimate owners of all firms and hence the ultimate bearers of underlying risks. And households hold the largest equity buffer, which in principle could increase the resilience of the financial system to adverse shocks. But they may not be as able as institutional investors to identify, judge and manage investment risks. In particular, it is unclear how effectively households will cope with financial market volatility. If households were to prove unable to manage the risks in DC pension schemes effectively, or if they set aside inadequate savings to provide for future pensions, their future retirement income would suffer. Thus appropriate regulation and supervision as well as consumer protection remain important, as do efforts to educate households on financial issues. In assessing influences on institutions' asset allocation decisions and on market dynamics, it is difficult to disentangle the effect of regulatory and accounting changes, which in most cases are either recent or still under discussion, from other factors. However, accounting and regulatory reforms do not appear to be a major cause of the current low levels of global longterm interest rates. Therefore, insofar as the behaviour of long-term rates may remain a puzzle, it is unlikely that a major part of the explanation will be found in the behavioural response of institutional investors to recent or prospective regulatory and accounting changes. Given the importance of assessing the equilibrium level of long-term interest rates, continuing uncertainty on this front complicates the task of policymakers. That said, developments in the United Kingdom, where long-term yields appear to have been affected by recent changes in institutions' asset allocation strategies, illustrate the potential importance of regulatory policy changes affecting institutional investors. In such specific circumstances, the response of long yields to regulatory and accounting changes will importantly depend on (i) the size of potential institutional demand for long-term bonds relative to the size of the market, (ii) the extent and scope of the regulatory changes, (iii) the initial funding or solvency positions of institutional investors and (iv) their initial asset allocations. Canadian real return bonds, all of which are issued at maturities of 30 years, may also have been significantly affected. Any assessment of the effects of regulatory and accounting changes on financial markets must be tentative, because the reforms are all recent or in the process of implementation or discussion. In addition, it is difficult to distinguish between changes that would have happened anyway (such as growing internationalisation of portfolios, asset-liability matching and shifts to alternative assets such as hedge funds and commodities) and those that may be a direct consequence of reforms. The fact that the regulatory and accounting reforms are part of a global phenomenon that is still in its early stages raises the possibility that they could eventually have a more lasting effect on financial market developments and prices. Policy implications The recent regulatory and accounting reforms seem likely on balance to enhance the functioning and stability of the financial system, and contribute to a more efficient allocation of resources. They should encourage better risk management by institutional investors, the spreading of investment risk among a larger investor base and improved transparency in corporate accounts. In the case of emerging markets, the growing demand from global institutional investors for emerging market assets is likely to be positive for these economies, and should contribute to the depth of local financial markets. The growing role of global investors in emerging markets might nevertheless alter the transmission mechanism of domestic monetary policy, especially if long-term bond yields become more dependent upon global factors. However, with the shift in the pensions sector from DB to DC plans and in the insurance sector from guaranteed to unit-linked products, the household sector has become increasingly exposed to financial markets, as the pooling of risk across time for workers disappears, and prospective retirement income is more subject to financial market volatility. This is also the case in emerging markets, where pension reforms are also encouraging the establishment of DC plans. While the reforms appear beneficial for financial stability in the long term, the implementation of these measures may temporarily distort prices in financial markets, eg through feedback effects with the potential to drive long-term interest rates below the levels justified by macro fundamentals. Therefore, during the transition to the new regimes, policymakers will need to take into account the risk of triggering unnecessary market volatility or distorted valuations. That said, the degree to which a gradual approach is possible will depend on the solvency situation, balance sheet risks and financial strength of affected companies, which may well vary from country to country. Finally, while the assessment is that these effects would be temporary in those cases where they exist, only after enough time has elapsed can one ascertain how long-lasting such effect would really be. These considerations highlight the need to conduct comprehensive impact studies prior to the completion of the reforms, in order to assess their effects and address any issues that arise. The increased interest in alternative investment strategies on the part of traditional institutional investors was not perceived as a major problem for financial stability, given that these asset classes still remain of limited importance in their portfolios and are mainly used to improve portfolio risk diversification. However, concerns have been expressed about the opacity of these vehicles, tail risks and untested markets. Regulatory policy changes may also increase the transparency of the existing links and channels of risk transfer between banks and institutional investors. In addition, they may help to provide increased transparency on the nature and location of the risks facing f inancial conglomerates, reducing the opacity that has existed in the past within complex financial institutions. Finally, the Working Group has encountered various limitations and challenges in using balance sheet data to study the investment behaviour of institutional investors. In articular, balance sheet data do not reflect accurately the risk exposures of institutional investors that are significant users of derivatives. If the CGFS wants to continue monitoring how pension funds and insurance companies are responding to regulatory policy changes, it needs to consider in more detail how to improve the information and/or analytical frameworks for assessing financial stability issues.