Chart 1: Consensus expectations of GDP growth in 2003 in the main markets, measured at different times through the year. Per cent

Sources: Consensus Economics Inc.

The low interest rate level fuelled strong growth in the US economy through the last half of the year. Sales of new and resale homes picked up, as did furniture, electronics and car sales.

Strong consumer demand in the US surprised many economists who expected that households would save a higher share of their income than was actually the case. Contributions to consumers’ income came primarily from government transfers and tax cuts. The business sector also took advantage of tax stimuli by increasing investment. Company investment also rose because of improved earnings and because some investment had been postponed after the stock market decline in 2001.

Strong demand for consumer and capital goods resulted in rapid destocking which in turn led to an increase in new manufacturing orders. A weaker dollar, especially in relation to the euro, also provided a stimulus to US manufacturing. Strong demand in the US did not lead to an increase in the current account deficit in general, however, although the deficit vis-a-vis China rose steadily. At year-end, the US had a larger trade deficit vis-à-vis China than vis-à-vis Japan. This is partly attributable to China’s development into a very competitive manufacturer. Another contributing factor is that China’s currency is tied to the US dollar. China is not only a significant exporter of goods to the rest of the world, especially the US, but is also an increasingly important source of demand for commodities and capital goods from the rest of the world.

Table 2: The market value of the Petroleum Fund’s sub-portfolios in 2003. In millions of NOK

GDP growth in Japan in 2003 is expected to exceed 2.5 per cent. Public investment and private consumption have not contributed substantially to growth. Developments in retail trade have been sluggish despite some improvement in the labour market. Household demand for credit has been weak and the savings ratio has fallen substantially in the last few years. This partly reflects a difficult labour market but also a rise in the number of pensioners. The Government has sought to change the policy of the 1990s whereby demand was stimulated by means of large public works projects. As a result, the public sector has not contributed to GDP growth in recent years. Growth in Japan has been tied to the export sector and to investment by companies that have experienced demand growth.

Growth in Europe has been mixed. The UK, Ireland and Spain are among the countries that have experienced the strongest growth, while Germany and Italy are among the countries with low growth. Steady demand for labour in the service sector has buoyed consumption in the UK, Ireland and Spain. Credit growth and demand have remained steady as a result of low interest rates in relation to wage growth and steadily increasing house prices. Consumers in Germany have been far more cautious, partly because unemployment has increased and partly because of the debate on reductions in various welfare benefits. A stronger euro, especially against the US dollar, has also put a damper on activity in many export companies.

Table 3: Transfers to the Petroleum Fund’s international portfolio in 2003. In billions of NOK

Many of the strong economic trends from 2002 and 2003 have continued. Household consumption, especially of consumer durables, has continued to rise in the US and the UK. In the US, this trend has continued despite a persistently difficult labour market in 2003. Lower interest rates, tax cuts and sharply increasing house prices have more than offset the negative impulses from the labour market. Companies have continued to reduce costs in 2003, especially labour costs. In spite of an increase in investment at the end of the year, companies are still cautious about new investment. The strong economic growth in Asia, and especially in China, has received far more attention than earlier from economists and investors. The size and rapid growth of Asian economies have global consequences for everything from commodity prices and freight rates to currency movements and inflation.

Table 4: The return on the Petroleum Fund’s ordinary portfolio for each quarter and for 2003 as a whole. Per cent

2.2. Return on the Petroleum Fund

At the end of 2003, the market value of the Petroleum Fund’s international portfolio was NOK 845.3 billion. This was an increase of NOK 236.3 billion since the beginning of the year. Besides the transfer of NOK 103.9 billion from the Ministry of Finance, the most important reasons for the increase were high returns on equity investments and appreciation of the currencies in which the Fund is invested in relation to the Norwegian krone. However, changes in the krone exchange rate have no effect on the Fund’s international purchasing power. Table 2 shows the size of the sub-portfolios at different times in 2003.

During the year, the Ministry of Finance transferred NOK 103.9 billion to the Fund’s krone account, and the equivalent of this capital was transferred immediately to the Fund’s portfolio of international securities. Table 3 shows that most of the transfers in the first part of the year were invested in the equity market, whereas all new capital since June 2003 has been invested in fixed income instruments. The distribution of transfers between the two sub-portfolios is designed to maintain the Fund’s equity and fixed income shares at 40 and 60 per cent respectively. As a result of this guideline, the Fund normally buys securities when prices have fallen.

Transfers through the year are based on an estimate of government net cash flow from petroleum activities. A correction will be made in the following year for the difference between the budget estimate and actual outcome. In 2002, the transfer was NOK 3.9 billion too high, and the Ministry of Finance took this into account when fixing the amounts to be transferred for 2003.

In 2003, the return on the Petroleum Fund’s ordinary portfolio (i.e. excluding the Environmental Fund) was 12.6 per cent, measured in terms of the currency basket which is defined by the country weights in the benchmark. In absolute figures, the return measured in terms of the currency basket was NOK 91.1 billion (excluding an increase in value of NOK 41.3 billion due to the depreciation of the krone). Table 4 shows that the return was negative in the first quarter, but positive in the last three quarters of the year. The fluctuations in the return figures generally followed developments in the most important equity markets.

The last column of Table 4 shows the difference between the actual return and the return on the benchmark as calculated by the index supplier. The return differential for the whole year was 0.47 percentage point. However, this difference does not provide a correct picture of the excess return that Norges Bank achieves by its management of the Fund. Even if Norges Bank were to follow the benchmark portfolio exactly (indexing), the actual return would not be identical to the benchmark return calculated by the index supplier.

In the management of the Petroleum Fund, extraordinary transaction costs accrue when new capital is to be invested in markets and when the Ministry of Finance changes the Fund’s benchmark portfolio. The Petroleum Fund also pays tax on share dividends in a number of countries. None of these cost components are deducted when the index supplier calculates the return on the benchmark portfolio. Due to the large transfers of new capital, extraordinary transaction costs and tax expense may comprise substantial amounts and can thus have a considerable impact on measurement of management performance. In Table 4, as in previous annual reports, adjustments have been made for tax on dividends and transaction costs in connection with changes in the fixed income benchmark. These components amounted to a total of 0.12 per cent of the total portfolio. After these adjustments, the gross excess return resulting from Norges Bank’s management of the Petroleum Fund in 2003 was 0.59 percentage point.

This gross excess return amounted to about NOK 3.6 billion. Of this amount, estimated extraordinary transaction costs associated with the fixed income portfolio are estimated at about NOK 421 million, and the tax costs associated with the equity portfolio at about NOK 339 million. Transaction costs in connection with the Ministry of Finance’s changes in the regional distribution of the equity benchmark and costs due to changes in the FTSE benchmark portfolio have not been taken into account. These costs are estimated at 0.14 per cent of the equity portfolio in 2003 and have accrued mainly in connection with internal equity management.

Table 5: Contributions to gross excess return in 2003. Percentage points

Table 5 shows that both equity and fixed income management made a positive contribution to the excess return in 2003. External managers made the most substantial contribution to the equity portfolio, while external and internal managers made very similar contributions to the fixed income portfolio.

The gross excess return is comparable with the excess return reported by other managers. However, it does not provide a true measure of Norges Bank’s net contribution to portfolio performance. The Petroleum Fund could have been managed passively, with a portfolio that was always kept very close to the benchmark. Instead, Norges Bank has chosen to engage in active management, which involves higher costs, but also yields higher returns. The value added by active management, which is calculated in Table 6, is a measure of the net contribution of this choice to the Fund’s return in 2003.

Table 6: Net value added by active management. Percentage points

Passive indexing involves a minimum of operating costs associated with management of the Petroleum Fund. Normal operating costs associated with indexing may be calculated on the basis of the costs of a group of comparable funds (see box in Section 7). For the Petroleum Fund, these normal management costs associated with indexing can be estimated at 0.04 per cent of the total portfolio. The estimate is a cautious one as it is lower than the median cost figure for the peer group. In 2003, total management costs amounted to 0.10 per cent, i.e. the costs associated with active management are estimated at 0.06 per cent.

With passive indexing, transaction costs would have accrued in addition to operating costs when the composition of the benchmark was changed. One of the world’s largest index managers, Barclays Global Investors, has calculated the normal annual transaction costs of following the equity index used by the Petroleum Fund at 0.07 per cent of the equity portfolio, or 0.03 per cent of the Fund’s total portfolio. The costs of following the fixed income index have not been calculated in the same way, but Norges Bank estimates them at 0.02 per cent, or 0.01 per cent of the Fund’s total portfolio. Thus, normal annual transaction costs of maintaining index management amount to about 0.04 per cent of the total portfolio.

On the other hand, with passive management there would also have been some revenues from lending the securities in the portfolio. In 2003, revenues from securities lending amounted to 0.04 per cent of the total portfolio, 0.03 percentage point of which could also have been achieved by passive management.

With these adjustment items, the net value added by active management in 2003 is calculated at 0.54 percentage point. This amounts to about NOK 3.2 billion.

Table 7: Return on the Petroleum Fund’s total portfolio in 2003 measured in various benchmark currencies. Per cent

Table 7 shows the return on the Petroleum Fund’s total portfolio in different currencies. Including the Environmental Fund, the return measured in terms of the currency basket was 12.59 per cent, whereas measured in NOK it was 19.96 per cent. The difference is due to a 6.54 per cent appreciation of the currency basket against the Norwegian krone in 2003. Calculated in euros, the return was 3.93 per cent, whereas the return in USD was 24.92 per cent. This is because USD depreciated against most other currencies in 2003.

Chart 2: Developments in the most important fixed markets in 2003. Yields on 10-year government securities. Per cent per year

Source: JP Morgan

Chart 2 shows that 10-yield yields exhibited a falling trend in the first part of the year but rose in the second half of the year.

Economists’ inflation forecasts for the main markets varied only slightly through 2003. Forecasts for the US varied from 2.1 per cent to 2.5 per cent, while forecasts for the euro area hovered around 2 per cent through the year. Forecasts for Japan ranged from -0.5 per cent to -0.8 per cent until the summer when forecasts edged up to -0.2 per cent at year-end. Actual inflation in the US fell from just over 2 per cent at the beginning of 2003 to just under 2 per cent at year-end. The rise in prices slowed for most goods and services in the US. The price decline was most pronounced for consumer durables, especially cars and computer equipment.This is because productivity growth has been very strong in these sectors of US manufacturing. Competition from imported goods is also intense in this part of the economy.

The rise in prices for services has also been falling, though not at the same rate as for consumer durables. Price inflation in the service sector is more closely tied to wage growth, since labour is the main input factor in products from this group. A weak labour market in recent years has contributed to low wage growth. A weaker dollar has not yet resulted in any substantial rise in import prices. This partly reflects that a large share of imports to the US come from Asian countries which have generally kept their currencies stable against the US dollar.

These Asian countries have a large trade surplus vis-à-vis the US. In addition, dollars have flowed into these countries, especially China, to finance a high level of direct investment. This in combination with currency interventions, where central banks have purchased dollars for local currency, has resulted in a considerable dollar surplus in Asian central banks. The central banks have largely purchased US government securities. Therefore, even with low yields, it has not been difficult to find buyers who are willing to finance the steadily increasing federal deficit in the US.

Chart 3 shows that returns on government bond investments in Asia/Oceania were negative during the year, whereas returns on comparable investments in the US and Europe were between 2 and 3 per cent.

Chart 3: Movements in Lehman Global Aggregate’s government securities indices in the main markets in 2003 (31.12.2002 = 100)

Source: Lehman Live

The difference between yields on credit securities and yields on government securities has narrowed sharply in 2003 (cf. Chart 4). This partly reflects companies’ substantial cost reductions in the last few years. Increased demand for companies’ products combined with lower costs has boosted corporate earnings. Higher earnings combined with a low level of investment have improved most companies’ cash flow. Increased cash flow means greater security for lenders and lower yield spreads in the market.Credit spreads have narrowed steadily through the year in spite of fluctuations in equity and fixed income markets.

Chart 4: The difference between yields on credit securities* and government securities (credit spread) in the US in 2003. Percentage points

* Companies with credit rating AAA from S&P Source: Lehman Live

Chart 5 shows the currency distribution of contributions to the fixed income return measured against the basket of currencies in which the Petroleum Fund is invested.The largest positive contribution came from investments in the euro countries. Contributions from the other currency areas were small. The negative contribution from investments in USD is due to the depreciation of this currency against the other currencies in the basket.

Chart 5: The individual countries’ contributions to fixed income returns measured in terms of the currency basket in 2003. Per cent

Chart 6 shows that the return on fixed income instruments in 2003 was slightly more than half of the average level prevailing over the last 24 years. Early in the 1980s, inflation rates and interest rates were unusually high. Both inflation rates and interest rates have been much lower in the last twenty years. Therefore, nominal returns on fixed income instruments must be expected to be lower on average in the period ahead than they were early in the period.

Chart 6: The return in fixed income markets 1980-2003 with the Petroleum Fund’s benchmark portfolio, measured in terms of the currency basket. Per cent per year

3.2 Management of fixed income investments

The market value of the Petroleum Fund’s fixed income portfolio rose in 2003 from NOK 378 billion to NOK 484 billion at year-end. NOK 64 billion in new capital was transferred to the portfolio during the year. The increase in value was a result of both positive returns in fixed income markets and a weaker Norwegian krone in relation to the investment currencies.

The inclusion of non-government-guaranteed bonds in the Petroleum Fund in 2003 has led to considerable changes in the composition of the fixed income portfolio. Great emphasis has been placed on phasing in non-government-guaranteed bonds in a cost-effective manner. Transaction costs have been systematically measured against estimated benchmark costs.

About 90 per cent of the fixed income portfolio has been managed internally in Norges Bank. There are two types of management. On the one hand, there is indexing and active management that is directly related to the indexing task. The objective of this enhanced indexing is to maintain a portfolio that is very close to the benchmark, while taking advantage of special pricing situations to achieve an excess return. In the fixed income portfolio, three sub-portfolios are indexed: government guaranteed bonds, fixed income instruments issued by private companies and collateralised securities. The three sub-portfolios are indexed internally, with the exception of securitised bonds in the US, which are managed externally.

On the other hand, there is more traditional active management, with positions that seek to take advantage of systematic price differences among fixed income instruments with almost identical properties, or with positions that will yield an excess return if yield curves or exchange rates move in a particular direction.

Table 8: Fixed income returns for each quarter and for the year 2003. Per cent

* The costs of reducing the weight of Asia/Oceania and phasing in non-government-guaranteed bonds
** The return differential is calculated as an arithmetic difference, i.e. in percentage points

The phasing-in process has resulted in extraordinary transaction costs. These costs apply especially to the purchase of fixed income instruments issued by private companies because each debt issue is considerably smaller than government debt issues and is thus less liquid. The extraordinary transaction costs relating to the total purchase of non-government-guaranteed-bonds through the entire phasing-in period has been estimated at approximately NOK 900 million, NOK 420 million of which applies to 2003.

At the end of 2003, the entire phasing-in process designed to achieve a distribution between government securities and non-government-guaranteed-bonds equivalent to market capitalisation weights had been completed.

Chart 7: Price performance in the FTSE equity indices in the major markets in 2003 (31 December 2002 = 100)

Source: FTSE

Chart 8: Price performance in the FTSE equity indices in the emerging markets in which the Petroleum Fund was invested in 2003, measured against the US dollar.

Source: FTSE

Chart 9: The FTSE All-World Equity Index. 1999-2003. Total and for the technology, media and telecommunications (TMT) sector. 31 December 1998 = 100

Source: FTSE

Chart 9 illustrates the significance of the technology, media and telecommunications (TMT) sectors for the equity indices over the past four years. Equity prices in the TMT sectors rose sharply from 1999 until early summer 2000 and then fell even more sharply. Developments in equity prices in the TMT sectors have been on a par with the other sectors in the FTSE global index in 2003.

Chart 10: Individual countries’ contributions to equity returns measured against the currency basket in 2003. Per cent

Chart 10 shows that in 2003 equity markets in all countries made a positive contribution to the Petroleum Fund’s equity return measured against the currency basket. The largest positive contribution came from the US, where the Petroleum Fund has the most substantial equity investments. However, weak developments in the dollar against other currencies in the basket reduced the contribution from the US.

Chart 11: Equity market returns 1980-2003 with the equity benchmark portfolio, measured against the Petroleum Fund’s currency basket. Per cent

Chart 11 shows that 2003 returns were good, but not extraordinary. The three previous years stand out in terms of negative returns.

4.2 The Fund’s equity investments

The market value of the Petroleum Fund’s equity portfolio rose in 2003 from NOK 231 billion to NOK 361 billion at year-end. The figures also include the environmental portfolio. The increase was due to outstanding returns in equity markets and the depreciation of the Norwegian krone in relation to the currencies in which the Fund is invested. NOK 40 billion in new capital was also transferred to the equity portfolio through the year.

In 2003, the portion of the equity portfolio under external management varied from 35 to 45 per cent. At year-end, 41 per cent was under external management. Regional mandates for active management accounted for most of this, although the portion representing specialist mandates for individual sectors is growing. External managers are responsible for approximately 66 per cent of the risk-taking in the equity portfolio.

Norges Bank Investment Management is responsible for indexing the equity portfolio. The purpose of index management is partly to reduce management costs and partly to better exploit opportunities for excess returns by managing the index portfolios actively (enhanced indexing). It is important to tailor the portion of the portfolio that is to be indexed and the portion that is to be managed actively.

At the end of 2003, NOK 213 billion of the equity portfolio was under internal management. Of this, 41.4 per cent was managed actively, while 55.6 per cent was indexed with cautious use of active strategies to take advantage of special pricing situations in the market to achieve excess returns (enhanced indexing). The remaining funds were in transition portfolios awaiting transfer to external managers. Overall, the scope of active management in 2003 increased from 57 per cent of the equity portfolio at the beginning of the year to 66.7 per cent at year-end.

Table 10 shows that in 2003 the return on the Petroleum Fund’s ordinary equity portfolio was 22.84 per cent, measured in terms of the Fund’s currency basket. The return was negative in the first quarter and positive in the last three quarters of the year. This reflects general price movements in equity markets through the year.

Table 10: The return on the Petroleum Fund’s ordinary equity portfolio for each quarter and for 2003 as a whole. Per cent

*The return differential is calculated as an arithmetic difference, i.e. in percentage points

The actual return on the equity portfolio alone was 0.69 percentage point, or 69 basis points higher than the adjusted benchmark return. The adjustment made was for tax on dividends received. About 75 per cent of the contribution to excess return was attributable to external management, while 25 per cent was attributable to internal management in Norges Bank. These figures do not take into account internal managers’ transaction costs in connection with restructuring portfolios for external managers. In addition, external managers are responsible for a considerable portion of the risk-taking, and as a result risk-adjusted excess return is higher for internal managers than for external managers.

In Table 10, adjustments have not been made for transaction costs in connection with the implementation of new regional weights in the benchmark portfolio or in connection with the inclusion of large and medium-sized companies in the benchmark portfolio. These costs are estimated at 0.14 per cent of the equity portfolio in 2003. The excess return includes income from equities lending amounting to NOK 205 million or 0.07 per cent of the equity portfolio.

In 2003, the return on the Environmental Fund was 22.89 per cent measured in terms of the Fund’s currency basket (see Table 11). This was 0.04 percentage point lower than the return on the Fund’s benchmark portfolio. The Environmental Fund is managed as a pure index portfolio, and the managers do not employ strategies to achieve excess return or enhanced indexing. The benchmark return on the Environmental Fund was 0.20 percentage point higher than the return on a comparable benchmark portfolio where companies had not been excluded on the basis of environmental criteria. This is largely because equity price performance in the TMT sectors, which constitute a large share of the Environmental Fund, was stronger than performance in traditional sectors (see the discussion of market developments above). Since the Environmental Fund was established in January 2001, the benchmark return has been 2.34 per cent lower than the return on an equivalent equity index where environmental criteria are not applied. This largely reflects the sectoral composition of the Fund’s portfolio.

Table 11: Return on the Environmental Fund in 2003. Per cent

Note to table 11: The table has been corrected 11 May 2004

Chart 12: Absolute market risk in the Petroleum Fund’s actual portfolio and in the benchmark at the end of each month in 2003. Per cent

Absolute market risk is chiefly determined by the benchmark portfolio defined by the Ministry of Finance. The Ministry has also set a limit for the tracking error associated with management, i.e. how much the actual portfolio can deviate from the benchmark. This tracking error shall always be less than 1.5 percentage points (150 basis points) (see box). Chart 13 shows that in 2003 the tracking error fell gradually from 40 to about 30 basis points for the total portfolio.

Chart 13: Expected tracking error at each month-end in 2003. In basis points (hundredths of a percentage point)

At the beginning of the year, the tracking error associated with equity management was considerably higher than the tracking error associated with fixed income management. The difference during the year narrowed, due in part to reduced volatility in equity markets. The higher tracking error in equity markets reflects that equity markets fluctuate more than fixed income markets, so that there is more risk associated with an active equity management position than with an active fixed income position of the same size. It also reflects that a larger share of the equity portfolio has been managed actively. Another measure of risk-taking in management is the share of the equity and fixed income portfolios consisting of securities that are not in the benchmark portfolio. The share of such securities in the equity portfolio remained stable at about 20 per cent in 2003, while the active share in the fixed income portfolio was lower.

Table 13: The fixed income portfolio as at 31 December 2003, by credit rating. Percentage of market value

Table 13 shows the composition of the bond portfolio (fixed income portfolio excluding cash) based on Moody’s and Standard and Poor’s (S&P) credit ratings. In the table, government securities and government-guaranteed bonds without credit ratings have been given the credit rating of the issuing country. According to the Ministry of Finance’s guidelines for credit risk, the Petroleum Fund may not normally be invested in securities with a credit rating lower than Baa from Moody’s, BBB from S&P or BBB from Fitch. Nevertheless, up to 0.5 per cent of the fixed income portfolio may be invested in securities with a rating of Ba from Moody’s, BB from S&P or BB from Fitch. All fixed income instruments have a credit rating from at least one of these agencies.

The fixed income portfolio contains long-term securities as well as fixed income instruments with shorter maturities, all of which have a credit rating of P-1 from Moody’s and A-1 from Standard & Poor’s.

Table 14 provides an overview of the risk limits stipulated in the Ministry of Finance’s Regulation on the Management of the Government Petroleum Fund and guidelines for the ordinary portfolio, and of actual exposure during the year. The figures show that exposure was within the limits at the end of each quarter.

Table 14: Risk exposure limits stipulated by the Petroleum Fund Regulation

Three breaches of guidelines laid down by the Ministry of Finance were reported in earlier quarterly reports for 2003. These breaches were committed by external managers and are not regarded as grave.

In the fourth quarter, it was discovered that one external manager had invested in a US unit investment trust as part of his cash management. The instrument is an Exchange Traded Fund and replicates the S&P 500 index in the same manner as equity futures. The investment raises questions regarding the interpretation of the guidelines for the Petroleum Fund. The Regulation on the Management of the Government Petroleum Fund stipulates that equity investments shall be made either in equity instruments or in derivatives of equity instruments. The investment in question was in units in a securities fund which, strictly speaking, is not an equity instrument. Although the instrument bears a strong resemblance to futures, it is not a derivative. Norges Bank chose to sell the holding in question.