Ten years of NBIM

In the spring of 1997, the Ministry of Finance decided that the Government Pension Fund – Global (then the Government Petroleum Fund) should be managed by Norges Bank. At the same time, the Ministry submitted a proposal to the Storting (Norwegian parliament) that parts of the portfolio be invested in equities. Just six months later, Norges Bank was ready to embark on one of the largest injections of capital into global equity markets of its time. During the course of a few hectic months in the first half of 1998, the Bank converted around 40 per cent of the bond portfolio into equities, a total of NOK 44 billion.

The mandate to manage the Government Pension Fund – Global was a vote of confidence in Norges Bank. But the mandate also presented considerable challenges. An internal working party at the Bank advised the Governor not to accept the mandate. The reasons for this were fear of a loss of reputation in the event of poor results, and the fact that this kind of activity fell outside the traditional role of the central bank.

Norges Bank’s management decided to establish a new investment management organisation, which was able to build its operations on existing expertise and systems at the Bank, in particular the Market Operations Department. The new investment management organisation was also given resources and opportunities to recruit from the Norwegian and global employment markets. A dedicated project was launched on 7 May 1997 to prepare for the first equity investments and serve as a forerunner for the new investment organisation. On 1 January 1998, NBIM was established as an operational investment manager.

This article takes a brief look at key milestones in the first ten years of NBIM. There is also a review of the results achieved for NBIM’s five core products: phasing in new capital, implementing the strategy set by the Ministry of Finance, creating value through active management, exercising ownership rights, and advising the Ministry of Finance on overall strategy. The article then presents some of the key choices made along the way which have led to the distinctive structure and culture which characterise NBIM as it is today.

One aspect of NBIM is perhaps more striking than any other. It is a publicly owned organisation within a central bank, but is, in every respect, run on commercial lines like a privatesector operator – indeed, perhaps even more closely guided by bottom-line performance than many private-sector operators. It is not often that this kind of culture is encountered in the public sector. So how did this come about?

The first ten years

Chart 1 shows movements in the number of employees and assets under management since 1998. At the end of the first year, NBIM had 41 employees managing a total of NOK 279 billion in three portfolios: the Government Pension Fund – Global, the investment portfolio in Norges Bank’s foreign exchange reserves, and the Government Petroleum Insurance Fund. Employee numbers grew sharply in 1999 when NBIM took over functions previously part of the Market Operations Department: securities settlement, IT and measurement of return.

 

Chart 2 shows the fund’s assets down between the four main types of management: external and internal, equities and fixed income. In the first year, external managers accounted for all equity management, and everything was invested in index strategies. Only in the latter part of 1998 were external mandates for active equity management awarded. In the summer of 1999, NBIM started up an internal equity management operation. At first, active management was given priority internally. Starting in 2000, however, index management was transferred from external to internal management. Since 2001, more than half of the capital in the portfolio has been managed internally. However, external managers have accounted for more than half of the risk taken in active management.

 

When it comes to bonds, Norges Bank itself handled all management itself to begin with. During the first few months, this was done by the Market Operations Department so that NBIM could concentrate on building up its equity management operation. From 2000, external managers were used to a certain extent. At first, this was solely for active management, but external index management was subsequently also included. This concerned the management of US mortgage-backed bonds, which were included in the management strategy with effect from 2002. In 2007, NBIM built up its own expertise in this field. External fixed income managers have invested around 10 per cent of the capital in the bond portfolio and have accounted for around 20 per cent of active fixed income management.

The current strategy plan for NBIM, approved by the Executive Board in January 2007, divides NBIM’s operations into five core products and sets specific targets for each of these. When looking back at the results achieved by NBIM during its first decade, it is natural to use this same subdivision into five core products:

- phasing new capital into the markets (transition)
- implementing the strategy adopted by the owners (beta management)
- excess return through active management (alpha management)
- exercise of ownership rights to safeguard the portfolios’ long-term financial interests
- advising the owners on investment management strategy.

Back in 1997 Norges Bank was considered as one of several potential operational managers for the Government Pension Fund – Global. However, the Government Petroleum Fund Act of 1990 stated that the fund’s capital was to be invested in the same way as the government’s other assets, which pointed towards Norges Bank, it being the government’s bank. At this time, the future size of the fund was uncertain. This, in itself, spoke against building up a separate institution. Norges Bank already had an expert environment for the management of its foreign exchange reserves in global bond markets and wellfunctioning systems for supporting and monitoring these investments. The Executive Board’s special professional responsibilities and independence are laid down in law and reinforced through practice. At the same time, investment management could prove a demanding role which could also affect the Bank’s reputation positively or negatively depending on how it was played. This role demanded the development of a clear commercial culture within an institution which is also to perform the traditional functions of a central bank.

 

Phasing in new capital (transition)

 Through the Government Pension Fund – Global, the government transfers its petroleum wealth into portfolios of global securities. Transaction costs play a significant role in the long-term net real return on the fund. From the outset, NBIM gave priority to developing its own expertise and systems to make trading in securities as efficient as possible. This section focuses on equity trading, which is generally far more expensive than trading in liquid bond markets.

In principle, trading is the final step in the investment process. Trading costs consist of commission, taxes, charges, market impact (bid-ask spread, price movements due to trading) and opportunity costs. Together, these costs can eat significantly into management results and make investment decisions potentially unprofitable.
There are four main types of equity trading at NBIM:

• Inflows of new capital. In 2007, these inflows averaged more than USD 300 million per trading day. The benchmark portfolio against which NBIM is measured has no trading costs.
• Maintaining the market portfolio (see following section on beta management). Once cash has been converted into equities, the portfolio needs to be maintained to reflect the return on the benchmark portfolio. Equities move in and out of the benchmark index for many reasons, and dividends need to be reinvested in equities. As with trading related to inflows of new capital, the cost of maintaining the market portfolio is not included in the benchmark for NBIM.
• Active internal management. As mentioned above, trading is normally the final step in the investment process. Investment ideas and decisions need to be executed in the market, and this is done by trading. NBIM’s equity team manages a substantial proportion of its assets internally, and this demands a fairly large number of transactions every year. In 2007, for example, internal trading for active strategies accounted for more than 60 per cent of total internal trading, or USD 250 billion.
• Active external management. Around 25 per cent of the equity portfolio is managed by external active managers. NBIM handles purchases of equities when establishing portfolios for external managers and when mandates are wound up (funding and defunding). Otherwise, the managers trade themselves. As with internal active management, the cost of trading in and out of the externally managed portfolios is potentially quite high.

 Chart 3 shows developments in quarterly equity trading volumes broken down into the main types of trade. Even if NBIM confined itself to index management, the trading costs associated with cash flows and indexing would be considerable. These costs are difficult to gauge, but we believe 25 basis points to be a fair estimate.

Since 1998, the equity team has consisted of two groups: one specialising in transition trading (cash flows and external manager funding and defunding), the other specialising in trading for internal index and active strategies (known as the single-stock group). Each group trades in all types of instrument (equities, futures, currency derivatives), and the two work together closely to build up systems and procedures to reduce total trading costs. Besides traders, the equity team has specialised operations groups whose role is to develop and maintain the necessary trading systems. NBIM’s equity department has systems, development, settlement and data groups which ensure the provision of systems which can support trading from start (initiation of a trade by a portfolio manager) to finish (settlement and safekeeping by the global custodian). An advanced trading infrastructure has been developed in order to interact with the marketplace (both trading and settlement).

Recent years have seen extensive changes in the structure for global equity trading. Two major changes which have impacted on the way in which NBIM organises equity trading are the growth in electronic trading and the availability of data.

Direct electronic trading on stock exchanges has exploded in the last five years – to the point where it is beginning to account for a substantial proportion of total trading for most trading operations on the buy side. Electronic trading has many potential benefits, such as speed, lower trading costs and anonymity. However, to be able to trade electronically efficiently and reap all these rewards, there is a need for traders with specialist expertise and for specialised and advanced systems. NBIM’s equity team began to look at electronic trading around the turn of the millennium and began building up systems early in 2003. Since then, electronic trading has come to account for a significant share of our trading and has contributed directly to reducing trading costs.

Chart 4 shows development in NBIM’s use of electronic trading for single stocks. Between 60 and 70 per cent are now traded directly on the stock exchanges.

The proliferation and availability of trading data due to advances in systems and architecture have given the equity team access to information and data which are critical to an understanding of the trading process and allow a reduction in total trading costs. What was difficult to calculate and analyse a decade ago is now available thanks to today’s systems. The large quantities of information and data available, on top of the number and size of trading transactions, provide constant opportunities to further develop trading systems and methods which can bring down total trading costs.

In the early years, NBIM’s equity trading was based in Oslo. NBIM has also had traders at its office in New York for the last four years, and in Shanghai since autumn 2007.

Chart 5 shows development in average equity trading costs at NBIM from 2003 to 2007. There is a clear downward trend, which can be put down to both electronic trading and other changes in NBIM’s ways of executing trades. Falling volatility in the markets is another important factor.

 

Market exposure (beta)

For large funds, expected risks and returns are determined largely by the overall choice of investment management strategy. Particularly important is the allocation between equities and bonds. In the case of the Government Pension Fund – Global, these choices are made by the Ministry of Finance after seeking the advice of Norges Bank and gaining approval from the Storting. The strategy is crystallised into a benchmark portfolio of equities and bonds.

One point of departure for NBIM’s investment management is investing all new capital in line with the benchmark portfolio. This kind of investment management is often known as index management. Another name is beta management. This name reflects how the aim of management is to achieve the exposure to systematic market risk determined by the client as its long-term strategy. The market risk priced in the capital market is often known in financial theory as beta. Excess returns beyond those resulting from diversified exposure to the market are known as alpha.

At NBIM, beta and alpha management are kept separate. There are several reasons for this. They are two entirely different types of management with different requirements in terms of expertise and systems. In both equity and fixed income management, separate groups have been built up specialising in beta management. This provides scope to realise economies of scale and focus on portfolio management and special techniques for identifying opportunities to generate slightly higher returns. Financial theory also provides grounds to separate beta and alpha in the operational implementation of investment management.

Beta management can be performed at very low cost. As demonstrated in a feature article published in 2002 pure index management of the Government Pension Fund – Global would not generate as high a return as the benchmark portfolio due to various types of transaction and management costs.

Beta management at NBIM has undergone major changes since 1998. This is due partly to the substantial growth in the size of the portfolios, and partly to changes in the strategies issued by the Ministry of Finance (and, in the case of the foreign exchange reserves, the Executive Board of Norges Bank). Chart 6 shows how these strategies have evolved over the last decade. By far the most important changes in the benchmark portfolio have been the inclusion of a 40 per cent allocation to equities in 1998 and the increase in this allocation to 60 per cent from 2007.

Chart 7 presents, as an example, developments in the Government Pension Fund – Global’s bond portfolio. It illustrates both the growth in volume and changes in its composition. To begin with, the portfolio was invested exclusively in government bonds. Bonds with credit risk were added from 2002, and it took around 18 months to complete the transition to the new benchmark portfolio. Inflation-linked bonds have also been added, and the allocation to US mortgage-backed bonds was increased from 2007. Until 2007, NBIM outsourced all index management of US mortgage-backed bonds to external managers due to the particular complexity associated with the option elements. As larger volumes increased, NBIM deemed it appropriate to build up its own expertise in index-managing these bonds.

New countries have been included in the equity portfolio, regional weightings have been adjusted, and, in the last year, small- and mid-cap companies have been explicitly included in the benchmark portfolio.

Index management is often referred to as passive management. However, there are constant changes to which the portfolio managers must adjust over and above the aforementioned changes in management strategy. New equities or bonds join the indices, while others leave. There are also events associated with individual stocks, such as the payment of dividends and the issue of their reinvestment, and changes in the weights of countries and regions (rebalancing). In NBIM’s index management, these changes and other openings are used in active strategies to try to achieve a slightly higher return than would result from pure index management. One important condition for this activity is very high-quality and up-to-date data, good IT systems, and portfolio managers capable of focusing on the detail.

In the early years, NBIM did not have sufficient equity management expertise in-house and so purchased index management services from external managers with specialist expertise. From 2000-2001, NBIM took over all such management. One reason for this was that NBIM wanted to use the index portfolio more actively in managing and implementing alpha management. In our equity management, much of the active management in the different subportfolios is broken down by sector and country or region. By keeping beta management inhouse, it became possible to fine-tune the allocation of external and internal portfolios. For example, we could then choose to index-manage in specific sectors where we did not see the same potential for active management as in other sectors.

In recent years, internal active equity management has been converted to a “long-short” system. Equities are borrowed from the beta portfolio, resulting in lower funding costs than from borrowing in the market. Internal beta management has become an important factor in being able to perform internal active management in the most efficient way possible.

The benchmark portfolio for bonds contains 9 750 securities. It would not be appropriate or cost-effective to buy all of these securities. NBIM’s beta portfolio contains around 4 000 securities. When picking securities, we attempt to reproduce the risk characteristics of the benchmark portfolio as efficiently as possible and to choose an overweight of bonds which can generate a slightly higher return. The risk limit for this type of active selection is far lower than for NBIM’s active management.

Excess return (alpha)

The return on the Government Pension Fund – Global is measured against the return on a benchmark portfolio and is calculated as the difference between the return on the fund and the return on the benchmark portfolio, and is a measure of how well the manager has executed the management mandate. This section looks at the results for the ten years from 1998 to 2007.(1

Chart 8 and 9 shows the excess return on the fund. The chart shows the annual and threeyear rolling excess return (2 on the fund as a whole and on the fixed income and equity portfolios separately.

During NBIM’s first decade, the average annual excess return on the Government Pension Fund – Global was 0.39 percentage points (39.2 basis points). The target in the strategy plans for NBIM has been an average annual excess return of more than 25 basis points over rolling threeyear periods (3. Chart 9 shows that the results have exceeded this by a good margin other than during a few months in late 2002 and early 2003. At the end of 2007, the average for the last three-year period was 29.6 basis points. During the last couple of months of this period, equity management accounted for the whole of the excess return generated, while fixed income management produced a negative average excess return for the first time.

Statistical analysis of returns

This section considers whether it is likely that the Government Pension Fund – Global’s excess return during the first ten years was due to chance or can more reasonably be attributed to management skill. Key properties of the excess return are analysed, such as the risk associated with active management, whether extremely high or extremely low excess returns have been more common, whether excess returns have had fatter tails than the normal distribution, and whether there has been any relationship between excess returns from one month to the next. The analysis links largely to Table 1, which presents the statistical properties of the excess return on the fund as a whole and the two asset classes: fixed income and equities.

The monthly excess returns on which the analysis is based are shown in Chart 10.

One of the limits set for the management of the fund is a ceiling of 1.5 per cent (150 basis points) for tracking error (relative risk, cf. Section 3.1.7 of the Annual Report), defined as the annualised standard deviation of the difference between the returns on the actual portfolio and the benchmark portfolio. This is a measure of the variation in the fund’s excess return. As can be seen from Table 1, the average standard deviation of the excess return (or tracking error) in the period 1998-2007 is estimated at 42 basis points.

The purpose of taking greater risks is to obtain higher expected returns. The information ratio (IR) is defined as the relationship between excess return and tracking error. The IR before costs for the period 1998-2007 is estimated at around 0.93. This is a relatively high IR, which indicates that NBIM has contributed high levels of value-added per unit of active risk (4. Statistical analysis makes it possible to assess whether this is due to management skill or chance. There are a total of 120 monthly observations, which provides a basis for calculating a test statistic (t-value) which, with given assumptions, follows a particular probability distribution. In this case, the t-value is an estimated 2.7. The probability of observing a t-value this high purely by chance is virtually zero, and we can therefore reject the hypothesis of the good results being down to chance.

Skewness

The empirical distribution of historical excess returns on the Government Pension Fund – Global does not show any significant skewness (see Table 1). However, closer analysis of equity and fixed income management individually reveals that the latter is significantly skewed to the left. This means that there are more low than high excess returns.

Fat tails

As illustrated in Chart 10, there are indications of fat tails in the empirical distribution. Fat tails mean that the probability of observing either high or low returns is greater than might be expected given normally distributed returns. This applies both to the fund as a whole and to equity and fixed income management individually (see Table 1). Whether excess returns are normally distributed or not can also be illustrated in a normal distribution plot (QQ plot). Chart 11 shows that there are more extremely high and extremely low excess returns than would be expected for normally distributed returns. This means that, when considering the fund’s relative risk, we must be careful about basing probability assessments on the normal distribution. It may be the case that the probability of observing both high and low excess returns is greater than the normal distribution would suggest.

Autocorrelation

It is an express aim of NBIM’s investment strategy to take a large number of investment decisions which are independent of one another both at a given point in time and over time. This investment philosophy has been presented in feature articles published in connection with previous Annual Reports. The philosophy is based on the Fundamental Law of Active Management (FLAM). When we analyse the fund’s return, it is therefore important to consider whether the return in one month is (linearly) independent of the return in the preceding month. The analyses show that it is not the case that an above-average excess return in one month is followed by an above-average return in the following months. Nor has any negative correlation between excess returns over time been identified. This independence over time in excess returns is consistent with NBIM’s investment philosophy.

The autocorrelation coefficients are shown in Chart 12. This correlation, which has been calculated for a time lag of 1-12 months, shows the relationship between the return in a given month and the return 1-12 months earlier. The solid blue lines on the chart indicate the 95 per cent confidence interval for the coefficients. None of the correlations fall outside this interval and are therefore significantly different from zero. The hypothesis of autocorrelation is also rejected by other tests (see Table 1).

Active management and the fund’s total risk

This section considers whether active management has impacted on the fund’s total risk, whether the fund’s excess return has been independent of the return on the benchmark portfolio, and whether parts of this excess return can be attributed to the fund having had greater market exposure than the benchmark portfolio.

         

Chart 13 shows that active management had a limited impact on the fund’s absolute risk in the period 1998-2007. The contribution to risk from active management is less than 1 per cent of the fund’s total risk.

Chart 14 shows that active management increased the fund’s return considerably more than its risk. The excess return associated with active management is around 8 per cent of the fund’s total return.

 

Chart 15 shows that there has been no significant relationship between excess return and the return on the benchmark portfolio. Therefore, the fund’s total volatility has not been affected particularly by active management. For all practical purposes, there is no difference between the standard deviations for the return on the fund and the return on the benchmark portfolio.

Both the fund’s low relative risk and the fact that the fund’s total risk has not been affected to any great extent by active management are a result of the management style. There is little relationship between excess returns on the equity and fixed income portfolios, and there is also little relationship between the excess returns on the individual mandates within the equity and fixed income portfolios.

Portfolio return, market return and excess return

It would appear from Chart 16 that the points plot along a straight line with a slope of 1. The relationship between the return on the benchmark portfolio and the return on the actual portfolio can be analysed using a regression model (see box)

.

The results of the regressions are summarised in Table 2. The fund’s beta relative to the benchmark portfolio is an estimated 1.0059. This is not significantly different from 1.00, which means that the fund has had approximately the same market exposure as the benchmark portfolio. In other words, NBIM has not chosen to operate with a level of systematic risk which departs notably from that of the benchmark portfolio. Table 2 shows an R-squared of 0.9975, which means, for all practical purposes, that all of the variation in the return on the actual portfolio can be explained by variations in the benchmark portfolio.

The return on the benchmark portfolio explains “all” of the return on the actual portfolio. The same applies to risk. As can be seen from Table 3, more than 99% of the total risk is associated with the benchmark portfolio. The reason why security-specific risk is slightly lower than the contribution to risk from active management in Chart 13 is that beta has, in practice, deviated from 1 (slightly higher).

Management skill can be assessed by calculating how large the return on the portfolio should have been given the level of market exposure accepted, and then calculating the difference between the actual return and this “riskadjusted” return. There are different approaches to what can be considered a relevant risk to adjust for when calculating risk-adjusted returns. One is to start from the classical capital asset pricing model (CAPM). The use of the CAPM requires extensive use of assumptions and does not give an entirely accurate picture.

There are other models (such as multifactor models) which capture more of the real risk factors in a portfolio than the CAPM. However, the CAPM is often preferred due to its simplicity. Note that the term “risk-adjusted” is used in this section for a single-factor model in line with the classical CAPM. The risk adjusted return in this context implies to calculate how large the return on the portfolio should have been given the market exposure accepted. We use the same regression analysis as presented in Table 2 to analyse these relationships. If the return on the portfolio is greater than the “risk-adjusted” return, the manager has generated a “risk-adjusted” excess return and so “outperformed the market”. The opposite is the case if the “riskadjusted” excess return is negative.

As can be seen from Table 2, the average annual excess return on the fund is an estimated 39.2 basis points. The “risk-adjusted” excess return (alpha) is an estimated 36.4 basis points, which is 2.8 basis points lower than the unadjusted excess return and significantly greater than zero (t-value of 2.7).

The fund’s alpha is 2.8 basis points lower than the average excess return. This can be attributed to the combination of a beta in excess of 1.00 and a positive average return on the benchmark portfolio during the period. The actual portfolio has featured marginally higher systematic risk than the benchmark portfolio, which is allowed for in the calculation of the fund’s alpha. These results are consistent with the conclusions drawn in the feature articles published in 2004 (“Results of six years of active management”) and in 2006. (“Analysis of Norges Bank’s results”).

As shown earlier, the beta deviations have had little impact on the excess return on the portfolios. Table 4 shows that these deviations have not affected the portfolios’ relative risk to any great extent either. For the fund as a whole, the risk associated with beta deviation is less than 1 per cent of the total relative risk. This corresponds to around 0.2 basis point, while the security- specific risk is estimated at 41.8 basis points. The fund’s total relative risk is therefore 42 basis points. For the fixed income portfolio, beta deviation accounts for around 0.4 per cent of the total relative risk, corresponding to 0.1 basis point out of a total of 34.5 basis points. In the equity portfolio, beta deviation accounts for 2.3 per cent of the total relative risk, corresponding to 2.0 basis points out of a total of 88.8 basis points.

Defined using a simplified model universe, more specifically the capital asset pricing model (CAPM), the regression analysis performed indicates that active management has resulted in a significant risk-adjusted excess return during the period. However, it is important to note that this single-factor model (CAPM) does not capture all relevant risk factors, and that it is very difficult to calculate in practice how much of the contribution from active management can be attributed to management skill, and how much to individual exposure to priced risk.

Summary

The analysis presented above indicates that NBIM’s active management of the Government Pension Fund – Global has made a statistically significant positive contribution to returns. A significant excess return has been generated in both equity and fixed income management. This excess return has been achieved despite limited risk exposure in active management. A positive excess return has been produced in both bull and bear markets, and can be put down to a large number of individual positions.

Exercise of ownership rights

The Ethical Guidelines issued by the Ministry of Finance in November 2004 made the exercise of ownership rights one of NBIM’s most important tasks. As an active owner with holdings in more than 7 000 companies, NBIM is to safeguard and build financial assets for future generations by promoting good corporate governance and high ethical, social and environmental standards at companies.

The Norwegian debate about ethical guidelines started up in spring 1997. When the Storting first discussed the proposals from Norges Bank and the Ministry of Finance for the inclusion of equities, ethical criteria were among the topics covered. Since then, separate ethical guidelines have been part of the work on the design of investment management. Norges Bank conducted studies of the consequences of various forms of ethical criteria in 1998 and 1999. From 2001 until 2004, NBIM managed a separate sub-portfolio based on defined environmental criteria. The Ministry of Finance gave the Bank a separate benchmark portfolio for this environmental fund.

Norges Bank participated in the committee which performed a detailed review of ethics in investment management in 2002 and 2003, and endorsed the committee’s main conclusions. The new Ethical Guidelines in 2004 gave the Bank increased responsibility for the active exercise of ownership rights. This mandate, how NBIM has implemented it to date, and plans for the future are discussed in detail in Section 4.1 of the Annual Report and in the other feature article.

When equity management started up in 1998, NBIM delegated the role of voting on behalf of Norges Bank to external managers. In 2003, NBIM began to vote for shares managed internally. In 2005, it took over all voting for externally managed portfolios as well. However, voting at general meetings is just one of many instruments used by NBIM to achieve its active ownership ambitions. It is increasingly engaging in direct dialogue with companies’ boards and management, and teaming up with other investors to increase its influence.

The Ethical Guidelines require Norges Bank’s exercise of ownership rights to take account of the extremely long time horizon for the Government Pension Fund – Global’s investments and their high degree of diversification. This obligates the Bank to promote the markets’ long-term sustainability.

Strategy

Under the terms of the Management Agreement between the Ministry of Finance and Norges Bank for the Government Pension Fund – Global, the Bank is to advise on the overall future strategy for investment management. The Bank also contributed actively as an adviser in the original establishment of how investment management should be organised. Among the key recommendations was drawing a clear distinction between political responsibility for setting risk appetite and overall strategy on the one hand, and operational management on the other. The first recommendation to invest in global equity markets was issued in April 1997.

Recommendations from Norges Bank are issued in formal letters from the Executive Board, all of which are made public. NBIM has participated actively in work on formulating these recommendations, but until 2007 it was a separate staff function outside NBIM which had the main responsibility for contributing to the formulation of the recommendations. Responsibility for the Bank’s recommendations on investment strategy was transferred to NBIM with effect from 1 January 2007.

The Bank’s recommendations on investment strategy have covered both the fundamentals, such as the choice of asset classes and their weights, and other areas, such as the exercise of ownership rights and guidelines for ethical investment management, risk limits, choice of benchmark indices, and rebalancing regime. From time to time, the Bank has also issued recommendations of a more technical nature and interpretations of the Ministry’s guidelines.

The box below presents some of the recommendations made by Norges Bank to the Ministry of Finance over the years.

Key choices that were made

Over the past decade, NBIM has developed some distinct character traits as an investment management organisation. These may possibly explain many of the results achieved and help to shed light on the differences between NBIM and other international funds and fund managers. This section looks in more detail at these traits and at some of the specific choices which have been important in the development of NBIM.

These key choices have included the vision for the organisation, the strategy for its operations, the principles for the management and development of the organisation, and a set of core values.

Vision

In May 1997, the Investment Management Project presented its business plan to the Bank’s management. This formulated the following vision for what was to become Norges Bank Investment Management: “NBIM is to build up expertise of the highest international standard in the management of investments in global markets and generate an excess return relative to the specified benchmark by making the best possible use of risk limits. Considerable importance is to be attached to risk management, control and reporting. NBIM as a department – and the individual employees at NBIM – are to inspire confidence in the outside world.”

This vision stated that NBIM was to meet the highest international standards. The reason for this was straightforward: the Government Petroleum Fund would be one of the world’s largest capital bases. The quality of investment management would be of considerable importance to the Norwegian economy over time. The expertise of the manager therefore needed to be right up there with the best internationally.

However, the first sentence of the vision also set out a more concrete ambition for the future organisation: adding value through active management. This by no means went without saying. It would be pointless investing extra resources in trying to “beat” the markets given prevailing financial theory which assumed efficient markets – and which also largely confirmed the likelihood of such efficiency in the most liquid markets. The best approach would be to manage highly diversified market portfolios as cheaply as possible (index management).

This was supported by overwhelming empirical research – only a few investment managers have managed to add value consistently over time. The economists at Norges Bank and the Ministry of Finance were well aware of this research.

Nevertheless, the Investment Management Project was explicit about its aspiration to be among the few managers to succeed with active management. Several reasons for this were presented to the Bank’s Executive Board and Supervisory Council during the rest of 1997. One was that there was no point having ambitions to achieve the highest international standards without the organisation having a concrete target to work towards. The target of excess returns was very concrete, it was ambitious, and whether or not it was achieved could be seen from measurable results over time.

It was also suggested that successful active management could result in considerable additional returns – in other words, that this was an important end in itself. Another argument was that risk management and control would be better if some active management was allowed than if the goal from the outset was only to replicate the benchmark return. There was an apparent paradox in that slightly more market risk through a risk limit for active management would reduce the overall investment management risk. The explanation was that concrete targets for creating value laid down as expectations for managers, departments and individual employees would result in each and every one of them having to be involved in the quality of the fundamental inputs into investment management: risk measurement, return measurement, efficiency in securities trading, data quality, settlement and so on.

The final part of the first sentence of the vision connected with the strategy for the organisation: excess returns were to be generated by making the best possible use of risk limits. We will return to this below.

The second and third sentences of the vision were also key to an operation charged with managing a nation’s financial wealth: high standards of risk management, control, reporting and integrity. In the subsequent operationalisation process where this vision was translated into concrete targets, these requirements became the foundation for NBIM’s operations: any excess return is of no value if operations are not founded on good control and management systems and implemented with absolute requirements for integrity for both the organisation as a whole and the individual employee.

The conclusion from the first ten years is that active management at NBIM has proved profitable, and that this ambition was important in building up professionalism and expertise. Maybe the scepticism of economists at Norges Bank and the Ministry of Finance in the early years was just what was needed to spur a structured and judicious investment strategy.

Strategy

The Investment Management Project’s business plan set out a strategy for the development of the organisation during its first few years. Since 1999, new strategy plans have been adopted every two years. These have been the Executive Board’s most important instrument in the governance of NBIM.

During the early years, the key strategic issues were how to get the best possible results from active management, and what kind of structure for investments would maximise the chances of achieving excess returns consistently over time.

As mentioned above, the focus from the very beginning was on making the best possible use of risk limits. A feature article published in 2000 examined this in more detail and presented the Fundamental Law of Active Management (FLAM) (5. The essence of this theory is that the expected information ratio can be maximised by spreading active management across as many independent positions as possible. Put another way, rather than large top-down positions, there should be many more decisions each accounting for less of the overall risk limit, and investment management should be organised in such a way that these decisions are as independent of one another as possible.

Several factors played a role in the choice of FLAM as the starting point for NBIM’s active management. NBIM’s core expertise provided a good seedbed for applying financial theory and approaching strategy in a structured, near-scientific manner. This was partially an “inheritance” from the investment management environment at Norges Bank and was cemented through the recruitment of key personnel in the early years. Academic background and skills were given greater priority at NBIM than in many other Norwegian and international investment management environments.

FLAM was also a natural choice given the unique situation faced by NBIM in 1998 with the high degree of scepticism about active management in the world around it. If the first few years produced negative results, this could result in an about-turn not only in the development of the investment management strategy but also in the principles for building and operating the organisation. In FLAM, NBIM found a sound basis for not allowing individual decisions to have a major effect on the overall management outcome.

The choice of FLAM was confirmed in each of the four strategy plans for NBIM approved by the Executive Board between 2000 and 2007. In a feature article published in 2004 the strategy was presented once again. The principles and priorities were the same as in the article from four years earlier referred to above, but updated in the light of a few years of experience.

This strategy has been pursued in both equity and fixed income management. Priority has been given to active management based on fundamental analyses and relative value strategies. This has also been reflected in the choice of external managers. Originally, NBIM had three investment management departments: Equities, Fixed Income and Tactical Asset Allocation. With the last of these, risk exposure was naturally concentrated in a small number of positions. The results were not satisfactory, and the department was closed after three years.

FLAM also provided extremely important guidance on how NBIM should be managed and organised. Investment strategy and organisational principles became two sides of the same coin. To ensure the greatest possible level of independence between active decisions, the following principles were adopted:

- there was to be the greatest possible degree of delegation of investment decisions to groups and individuals - there was to be no overall investment stance or top-down management of investments - there were to be no committee decisions - both external and internal managers were to be used - when choosing external managers, managers with fundamental strategies were preferred - in internal management, the individual portfolio manager was to be given the resources and opportunities to build up specialist expertise

Outsourcing is another important strategic choice made by NBIM. From the very first strategy plan approved by the Executive Board, one guiding principle for NBIM has been to outsource as much as possible outside its core business. The reason for this is partly to focus limited management resources on tasks which it would not be possible to outsource, and partly to promote higher quality and more rapid development. The reason is not to cut costs. It was acknowledged from the very outset – and has been confirmed along the way – that outsourcing can result in slightly higher operating costs than in-house production. The profitability of choosing to outsource stems from better quality and better management of the remaining business.

Leadership

Even when it was first established, it was realised that NBIM would face continuous change over many years. Rapid growth in assets under management would also mean growth in the organisation and constant redevelopment of systems and investment management. It was also expected that the Ministry of Finance and the Storting would make changes in the overall investment management strategy.

These unique challenges were naturally associated with requirements for NBIM’s leadership to be change-oriented and flexible. No manager could count on retaining his or her position and responsibilities indefinitely: managerial duties were defined as a relay where, sooner or later, it would be time to pass on the batten. Management evaluations were performed and external help brought in to train both the senior management team at NBIM and individual managers.

From the beginnin NBIM was a small organisation with no need for numerous management tiers and with small formal differences between management and other employees. As with modern knowledge companies, management was based more on natural authority than formal authority. Technical expertise was, in practice, put on an equal footing with managerial competence. Together with the constant growth in duties and the associated increase in interesting career opportunities, this has made it easier to alter the management and structure of NBIM than at many other types of organisation.

The informal management style and cautious use of formal authority have persisted even as NBIM has grown in size and complexity. NBIM is not only a knowledge organisation – it is also dependent on recruiting, developing and retaining special talents. Experience from the first decade – and in the global investment management industry as a whole – is that only organisations with particularly skilled and talented people are able to add value for their clients. This makes unique requirements of management.

Perhaps the most important has been allowing people to do their jobs their own way without unnecessary intervention. Strong individuals need room to manoeuvre. But passive acceptance of this is not enough – employees need to be encouraged to take responsibility and to take control of how they will achieve the targets set. Talented key personnel also ask a lot of their surroundings – they want to have the best possible access to resources and systems, and cannot accept less being asked of others in the organisation than is asked of them themselves.

In a feature article on experience of the use of external equity managers published in 2004 this was summarised as follows: “One lesson that we have learned is that some individuals have a very substantial impact on results. Of course, very few individuals perform well in an organisational vacuum. Organisations perform best when a skilled portfolio manager is faced with a challenging investment environment.”

At NBIM, these management principles have coincided with and supported the strategy for investment management. To achieve the greatest possible level of independence for decisions on active investments, these must be delegated to many different groups and individuals. This delegation must be real, without intervention from superiors, provided that the individual employee stays within the agreed structure and risk limits. To ensure the best possible outcome for each individual investment decision, each individual employee must have an opportunity to develop his or her specialist expertise. The sense of ownership which derives from responsibility and an absence of intervention from superiors is an important motivation and driving force for the development of this expertise.

Structure ad systems

From the outset, NBIM has adhered to the principles of line management. For every krone in the portfolios, which are spread across more than 40 000 positions,(6 there is an employee at NBIM who is its “owner”. This also applies to portfolios managed by external managers. The individuals at NBIM responsible for choosing external managers are also responsible for the results they achieve. There are no committees along the line between the individual position in the portfolios and the Executive Director of NBIM. There has been a conscious decision to avoid line responsibility in investment decisions being muddied by group or committee decisionmaking.

The pay system has played an important role in clarifying responsibilities and encouraging employees to focus on the targets which management decides are important. Variable pay based on performance has been an important part of the system and has contributed to NBIM having succeeded to a great extent in retaining its most talented employees. (Section 5 of the Annual Report looks at NBIM’s pay system in more detail.)

• A simple description of NBIM’s structure might be as follows:
•  Roles and objectives are defined by the clientStrategy plans approved by the Executive
• Board provide further detail and guidance.s
• NBIM’s senior management expands on thiswith annual targets and action plans.
• This ends up in targets and plans for each individual employee.
• A wide range of management and control systems serve as resources in work on achieving the targets and contribute to independent control and follow-up. 
• For the individual employee, performance relative to these targets has consequences in terms of career opportunities, risk limits and remuneration.

Culture and values

The vision was very ambitious, and NBIM’s culture was soon characterised by a businesslike approach and a focus on tangible results. But the environment was also relatively academic and detail-oriented. One illustration of this is how, from the very outset, NBIM performed daily measurements of relative market risk across the portfolio. Resources were channelled into this ahead of the development of investment management, although hindsight has shown that daily measurements at an aggregate level are of rather limited value.

This example also illustrates another of NBIM’s character traits: the very great importance attached to management and controls. Activities are not launched until all conceivable checks have been carried out. In this respect, the organisation clearly shows its kinship with the central bank. The same applies to the importance attached to requirements for personal integrity and honesty. But there is less evidence of this kinship in the emphasis on a commercial approach and line management and on personal responsibility for measurable results.

NBIM has extensive contact with private investment managers and other large private and public institutional investors. This contact serves as a source of inspiration for the further development of NBIM’s organisation, but also as a good starting point for comparisons. One impression is that NBIM goes further in terms of line management, delegation and prioritising personal responsibility than even many private managers. Perhaps this has to do with our origins in a central bank. It was very clear right from the outset that the central bank and NBIM needed to have very difficult cultures to reflect the differences in their roles.(7

As NBIM grew, a need gradually

also emerged to formulate and follow up a number of core values to support the objectives for its operations and its culture. These were developed in 2004 after an extensive process involving all employees. The four core values – Excellence, Innovation, Integrity and Team Spirit – are discussed in more detail in Section XX of the Annual Report.

At NBIM, there is a clear connection between investment strategy (FLAM), line management with a high degree of delegation, and the emphasis on developing values and culture. Delegation is essential for implementing the investment strategy. But this high level of delegation can be achieved only when we can be sure that employees share our underlying values. NBIM follows up its core values in regular group exercises, in employee appraisals, and in the recruitment of new employees.

Continuous changes

The markets in which NBIM operates are constantly evolving. What might have been a good way of managing investments for some years may be unsuitable in subsequent years. The capacity for change is therefore an important success factor.

NBIM’s first decade can be seen as a process of continuous change. The growth in assets under management and changes in the owners’ investment management strategy have also provided important impulses for change. The structure of the organisation has evolved constantly, with consequences for many managers and other employees. In most cases, these changes have resulted in new career opportunities and been welcomed.

A major test faced by NBIM in the coming decade is maintaining its ability to add value through active management at the same time as assets under management may grow very rapidly. Significantly more assets under management are not a problem per se, provided that they are invested in large and liquid markets. NBIM has the expertise and apparatus to handle further large inflows of new capital and invest them efficiently with broad exposure to the markets.

The challenges lie partly in implementing active management and partly in investing substantial sums in less liquid markets. Some parts of active management cannot be scaled up as assets under management grow (see the feature article on investment strategy published in 2004). One approach would be to have larger individual positions, but this could also result in a slightly lower return for the active risk taken.

All comparable large funds are far more diversified across asset classes than the Government Pension Fund – Global. Common to the asset classes in which the fund is not yet invested – such as real estate, infrastructure and private equity – is that they are far less liquid than today’s equity and fixed income investments. Investments in these markets will bring new demands for specialist expertise and require extensive resources.

With more assets under management, new types of active investment strategy, and maybe also new asset classes, NBIM faces continuous challenges in recruitment, skills development and leadership. The ability and willingness to handle change will probably be as important in the second decade as in the first. Much of the growth in the organisation will probably need to be outside Norway, and NBIM will increasingly become an international investment management organisation.

More assets under management will give NBIM greater influence as an active owner. In implementing the Ethical Guidelines, it will be a particular challenge to ensure that NBIM remains an investment operation and is not charged with playing a purely political role. Whatever the size of the Government Pension Fund – Global, active ownership will only be effective for as long as companies and other investors view NBIM as a professional investor. At the same time, the fund’s long time horizon means that NBIM must engage companies on issues beyond just short-term earnings. Account needs to be taken of the sustainability of marketplaces and of external impacts, including ethical and social considerations. However, the approach to active ownership will be based on the instruments of an investor and not on political means.

1) The data and methodology upon which this article is based, are the same as those presented in the feature article published in 2006 where eight years of data were reviewed. The performance data are the same as those published in the annual report, but due to different calculation methodology there are some minor differences on an annualised basis.

2) Based on the monthly arithmetical average.

3) As discussed in Section 3.1.2 of the Annual Report, the target is more accurately for net value-added from active management. The chart shows only the gross excess return.

4) The information ratio provides an indication of thequality of active management has produced, adjustedfor the risk taken. However, one must be careful touse the ratio only as a guide to active managementperformance. There is no official standard in thefinancial sector for what constitutes a good/highinformation ratio. Nevertheless, information ratiosabove 0.5 are often referred to as very good

5) FLAM was launched a few years earlier by Grinoldand Kahn in their book “Active Portfolio Management”,Irwin Professional Publishing, 1995. 

6) A position is a manager’s investment in a security. If, for example, three managers all invest in the samesecurity, this counts as three positions.

7) Here is a stylised example. When it comes to monetary policy, a central bank takes only a few rate-setting decisions each year, which have significant consequences. Each decision must be subjected to extensive quality assurance. In an investment organisation like NBIM, thousands of decisions are taken each year, many of which will necessarily be incorrect. An investor is successful if just over half of his decisions are correct (given symmetry in terms of gains and losses). One of the biggest mistakes that can be made is not to be prepared to take risks when there is actually an opportunity to build an organisation with sufficient expertise.