Please note that the text below may differ slightly from the actual presentation.

Thank you for this opportunity to discuss the challenges that arise when managing financial wealth in a public sector environment. Norges Bank now has eight to ten years of experience of this task, and I shall try to describe in brief the strategies adopted by Norges Bank, and what we believe we have learned.

We have a large amount of money under management. Our foreign reserves are probably larger than needed for financial stability and monetary policy operations. But the most important fund is the Government Pension Fund, formerly the Petroleum Fund. We have been responsible for managing that Fund since its inception in 1996, and it has been a major task for us since 1998.

The value of the petroleum resources in the North Sea is currently estimated at about USD 700 billion. The government receives the bulk of the revenues generated from petroleum extraction as sales revenues and taxes. The government channels these revenues into the Pension Fund as financial assets.

Annual spending from the Fund is subject to a spending rule, whereby the government commits to limit spending. Over time spending should not exceed the estimated expected real return on the Fund's investments, set at 4 per cent per year. This implies that the Fund is expected to become a permanent savings account.

At the outset, there was already an investment management operation in the Bank, for the foreign reserves. But when it became clear in 1997 that the Pension Fund would be growing fast, we decided to set up a separate management unit, now called Norges Bank Investment Management (NBIM). Our challenge was to build a business organisation within a traditional civil service culture.

The new unit recruited investment professionals from outside the Bank as well as people from the existing reserve management operations. NBIM was explicitly defined as a business unit, which should be different from the rest of the central bank in many respects.

To avoid any suspicion that NBIM would benefit from proprietary central bank information, the new unit was placed completely outside the decision-making processes in the monetary policy and financial stability area, with no access to confidential information.

The fact that the Pension Fund does not invest in Norwegian markets is also important to avoid conflicts of interest: There is no direct link to our monetary policy operations.

The purposes of the Pension Fund and the foreign reserves are similar; they are both essentially national savings funds, except for that part of the reserves that we may need for policy purposes. The current asset allocation is thus effectively the same for both funds. At present they have 40 per cent in equities and 60 per cent in fixed income. The investments are widely diversified across international markets.

The investment strategy has so far aimed at low return volatility, perhaps to a larger extent than is reasonable for a fund that is meant to be permanent. The strategy is currently under review by the Ministry of Finance, and there may be changes under way.

The Ministry of Finance acts as owner of the Pension Fund, on behalf of the government. The management of the Pension Fund is delegated to Norges Bank. The key feature of the model is the division of responsibilities between the two institutions. The Ministry takes the high-level decisions and defines a clear mandate for Norges Bank. In accordance with that mandate Norges Bank makes all investment decisions without interference from the Ministry.

The benchmark portfolio for the Fund is defined by the Ministry. The benchmark determines most of the asset allocation in the actual portfolio, but the Ministry also leaves some leeway for the manager. NBIM shall deviate from the benchmark in order to obtain a higher return.

Active management has become a very important part of our investment operations. We believe that the search for excess return has had a strong disciplinary effect on the organisation with a view to controlling both market and operational risk, and to ensure that all inside and outside suppliers do exactly what they are paid to do.

We have tried to build an organisation where the focus is on investment returns. In order to keep that focus we have tried to implement three basic principles:

• Accountability for decisions is clearly defined
• There is transparency, nothing can be hidden away
• The professional standards required for employees are very high

For investment decisions these principles mean that every decision is made by one individual and that the remuneration paid to him or her depends on the profitability of the decisions. We have defined a large number of independent profit centers, each with a clearly defined mandate. A compliance function checks that no one operates outside the mandate.

For external managers we apply the same principle: Hiring an external manager is an investment decision, and someone inside NBIM is held responsible for it.

The salary system is different from the rest of the central bank. The main objectives are to focus attention on performance, and to give distinct signals to people who do not perform according to expectations.

Salaries have two components, a basic salary and a performance-based component. The latter is a very important part of the total salary, and implies that differences between the salaries of poor and outstanding performers are large. The level of salaries is benchmarked to private sector competitors in the local market where we operate.

The criteria used are designed to be transparent with little room for subjective judgement. Measured excess return contribution is a very important component.

I believe that our experience so far has been positive. Since the Fund's first full year in 1997, the annual return in international currencies has been more than 6 per cent in nominal terms and 4.1 per cent in real terms.

Since NBIM was set up in 1998, we have recorded the performance in 34 quarters. In 9 quarters the return to the Fund has been lower than the benchmark, but in 25 out of the 34 there has been a positive excess return. On average, NBIM has earned exactly half a percentage point higher return per year than the benchmark portfolio. In nominal terms, that adds up to somewhere between USD 4 and 5 billion in extra earnings since 1998.

In this introduction I have touched on what I believe are important factors behind the performance of NBIM. It is important to keep in mind that the purpose of investment management operations is to make money, so we need to have a business organisation. The responsibilities of each individual must be clearly defined, and his or her remuneration must depend on performance. The incentives must be set up in a way that leads each individual to chase the correct targets, and in a way that rewards superior performance.

This is not a description of a traditional central bank culture, so the investment management unit may need to be built outside the existing structures. At the same time the existing structures may benefit from having a business unit within the central bank. There are lessons to be learned about efficiency in operations, and that has also proved useful for the civil service part of Norges Bank.