Phase-in of new equity share for the Government Pension Fund Global
It was decided in 2017 to increase the equity share in the fund’s benchmark index to 70 percent. The Ministry issued a plan for phasing in the new equity share in its letter of 22 September 2017. This plan was based on advice from Norges Bank. At the same time, the Bank was asked to report on the financial consequences of the phase-in once it was completed. The Bank’s report follows below.
The phase-in of a higher equity share was implemented by gradually increasing the equity share in the strategic benchmark index over a period of 20 months from September 2017 to April 2019. In addition, a rule was set for how the equity share in the actual benchmark index should be adjusted should it fall below the equity share in the strategic benchmark index. It is only when the equity share in the actual benchmark index is adjusted that there is a need to perform trades in the portfolio.
When the phase-in began in September 2017, the equity share in the actual benchmark index was 66.2 percent. This was much higher than the equity share in the strategic benchmark index, which was then 62.5 percent. Equity prices rose substantially during the early months of the phase-in. The gap between the equity share in the actual and strategic benchmark indices was therefore maintained despite the strategic equity share being adjusted at the end of each month. After the bull market ended in early 2018, the gap gradually narrowed.
In October 2018, the equity share in the actual benchmark index fell below that in the strategic benchmark index for the first time. In line with the phase-in plan, the equity share in the actual benchmark index was therefore adjusted towards that in the strategic benchmark index. The equity share in the actual benchmark index was below that in the strategic benchmark index for the next three months and again in March 2019. The actual benchmark index was adjusted upwards during this period by a total of 2.9 percentage points. The fund’s average ownership share, measured as its share of the float-adjusted capitalisation of the benchmark index, increased as a result from 1.34 to 1.41 percent.
Purchases and sales of securities cause the portfolio to incur transaction costs. These costs will normally be higher when trades are made over short periods. The phase-in plan made allowance for this by spreading the process over a longer period, cf. Norges Bank’s letter of 21 August 2017. In addition, a ceiling was set for how far the actual benchmark index could be adjusted in a single month. The change in the equity share resulted in adjustments to the benchmark indices for equities and bonds amounting in total to almost 500 billion kroner. We estimate the transaction costs for these adjustments at around 630 million kroner, breaking down into 150 million kroner for sales of bonds and 480 million kroner for purchases of equities. We should stress that these costs are model-based.
The equity share in the actual benchmark index was adjusted on the last trading day of each month. To reduce transaction costs, the actual trades were carried out over a number of days. This timing discrepancy may have impacted on the fund’s relative return. Other investment activity proceeded as normal during the phase-in period. Actual trades were minimised by offsetting trades related to the phase-in against other trades. Actual transaction costs may have been both lower and higher than the model-based costs.
Consequences for the fund’s return and risk
A higher expected return on equities than on bonds in the long term speaks in favour of a rapid phase-in of a higher equity share. On the other hand, rapid adjustment means somewhat higher transaction costs and a greater risk of the timing subsequently proving unfavourable. This risk can be reduced by spreading the phase-in over time. The Ministry singled out costs and risk as the main reasons for setting a relatively long phase-in period.
In our opinion, the plan for phasing in the higher equity share worked well. Actual purchases of equities were carried out largely on the back of a period of falling share prices. However, substantial price movements during the period served to increase the complexity of the phase-in and had a negative impact on the fund’s relative return of around 2 basis points, or around 2 billion kroner.
Jon Nicolaisen Yngve Slyngstad