Petroleum Fund's active indexing
Last year Norges Bank decided to transfer substantial equity index portfolios of the petroleum fund from external to internal management. The main purpose of moving the portfolios was to use various techniques to achieve somewhat higher returns than the index – in other words, enhanced indexing.
Up to the end of 1998, the entire petroleum fund portfolio was managed by index managers. The international trend for several years has been to increase the portion of index management, not least because of the favourable results achieved relative to active portfolio management. Index management is a low-cost form of management and it provides a wide risk diversification for the portfolios.
As index management has become increasingly widespread, there is a risk that equity prices are influenced in an unfavourable direction for those who use an index strategy. The negative effect may primarily stem from price movements when companies enter or leave indices, or when outstanding equity capital is changed. In the FTSE index, which is the index that the Ministry of Finance has defined as the benchmark index for the petroleum fund, such changes may occur as many as 500 times in a normal year. One of the main reasons that Norges Bank prefers a new strategy for index management is in effect to avoid return losses as result of index changes.
Norges Bank started its internal index management in January 2000. At the end of 2001, the internal index portfolio amounted to a NOK50bn (E6.8bn) in the Petroleum Fund, and NOK23bn in the long-term portion of foreign exchange reserves.
With an equity index portfolio of this magnitude, it is more cost-efficient to engage in internal management. There are clear economies of scale within management in general and indexing in particular because the administrative work involved in portfolio maintenance is independent of the size of the portfolio. The cost of managing an additional capital unit is relatively small, and average costs are thus reduced.
Another merit of internal indexing is that it facilitates active management because the monitoring of corporate events, index changes and focus on data quality can be left to the index mandate/group. Internal indexing also provides greater scope for flexible management of the active management resources of selected regions or sectors.
However, the most important reason for choosing a large degree of internal indexing is the associated opportunities for achieving excess returns by using relatively simple strategies. This type of management is referred to as ‘enhanced indexing’.
Active indexing results in outperformance of the benchmark is what distinguishes this strategy from traditional passive indexing where the goal is to track the benchmark as closely as possible. The opportunities for enhanced index management are created by technical or structural aspects of equities or equity markets, and they exist for short periods of time. The strategies that are used are relatively precisely defined, and are implemented when the conditions are appropriate. Market risk is limited and is considerably lower for index management than for enhanced management. The expected excess return is also lower.
Changes to indices create opportunities for enhanced index management. When the weight of a company is changed or the company is entering or exiting an index, investors, particularly passive index funds, tend to implement the change at the index effective date. This behaviour creates price effects because of higher supply of or increased demand for the equities. One strategy is to implement changes at times other than the benchmark index effective date, and function as liquidity supplier of the equities at the time of a shift in other indices than our own index.
In 2001, there were substantial changes in the indices supplied by the two major index suppliers FTSE and MSCI. A common objective of the changes made by the two index suppliers was to achieve a better accord between the supply of marketable equities, so-called ‘free float’, and the weight of these equities in the indices.
By taking over direct control of these large portfolios it became easier to implement the necessary changes in the portfolios – well distributed over time and in due time ahead of changes in the indices themselves. Some of the portfolio shifts were made by channelling transfers of new capital for the petroleum fund to the large index portfolios.
Early adaptations to index changes that we knew would occur, entailed deviations from the benchmark index up to June 15. These positions were taken because we believed that we could achieve higher returns than if we had waited for the actual index changes to be implemented. The purpose here was to act before others made the same shift. This strategy involved a true risk, and in order to reduce the risk we applied hedging strategies both in the equity market, where we used co-varying equities in the same industry, and in the market for equity futures at country level.
A considerable portion of the index group’s performance contribution in 2000 stems from this early adaptation to the substantial changes in the index, which enabled it to achieve excess returns.
Norges Bank uses one external manager for enhanced indexing. The strategies that are employed are, in principle, the same as for internal index management, but the focus may vary somewhat. It is an advantage to have an external manager whose performance can be compared with our own – a dynamic benchmark for comparing internal and external index management.
The external manager has long experience in the field of such management, and we have benefited greatly from the professional contact.
Enhanced indexing involves applying strategies to various phenomena that have been tested in empirical research. The strategies focus on broadly observed phenomena and structural changes linked to specific equities and equity markets in general rather than to in-depth/fundamental evaluations of individual companies. The strategies are still implemented as positions in individual companies. The empirical testing of the strategies show that as a rule the excess returns are significantly positive in relation to a benchmark.
2001 was a year that offered good opportunities within enhanced indexing – particularly in connection with the considerable changes in the FTSE and MSCI indices. The index strategies generated an excess return of 0.35 percentage point of indexed capital in 2001. The main contribution to the excess return came from the positioning in connection with the changes in the FTSE index. This shift turned out approximately as expected, with some surplus demand for (supply of) equities whose weight was increased (reduced) in the index.
As to the shifts in the MSCI index, there was an effect opposite to what we expected the last two days before the shift was implemented at the end of November. This shows that investment strategies are far from risk-free.
The results can be characterised as very favourable by the measures normally used for this type of management, given the sizeable amount of capital under management. Internal index management generated a somewhat higher return than external index management in 2001.
The opportunities for relative excess returns in enhanced indexing are, however, reduced when the capital under management increases. This is because there are limits to the size of the positions that can be taken on each occasion. The results achieved using this type of management will vary in relation to the investment opportunities that arise and how large a transaction volume it is then possible to have traded at advantageous prices.
The total excess return on indexing (0.35 percentage point) is higher than both the internal (0.33 percentage point) and the external (0.26 percentage point) rate of return. This is because internal and external returns are calculated monthly in relation to the capital that the internal/external managers have managed.
This article has been written by the Norges Bank Investment Management team responsible for the Petroleum Fund in Oslo