What is the optimal allocation towards real estate in the portfolio of the global pension fund global?
MetadataVis full innførsel
- Master Thesis 
In this master thesis we evaluate the optimal future investment allocation towards real estate for the Norwegian Government Pension Fund Global (GPFG). Based on an assessment of the relative risk and return attributes of equities, bonds and real estate -‐ and using a mean-‐variance optimization – we have found that the fund should allocate a full 11,2 % of its capital towards real estate (59,4 % to equities and 29,4 % to bonds). This is twice the current target level, and would represent an additional 235,6 BNOK (42 BUSD) of GPFG funds being allocated to investments in the global real estate markets. In performing the above analysis we have been able to rely on a fairly well documented analysis based on long term global data for the performance and volatility of bonds and equities. Our key focus has been to assess and derive the appropriate performance characteristics of real estate. By doing looking at different property data, we have been able to develop a well-‐founded view of the historic performance of real estate over the last 25 years. In addition to this we have made a qualitative assessment of the asset class and have used this to develop what we feel are robust and reasonably conservative estimates for the expected future performance characteristics of a global property portfolio. Because of several specific characteristics of real estate it has been argued that it cannot be analyzed in a simple mean variance framework. We have therefore tested the robustness of our findings by applying additional perspectives and approaches. On this basis we remain convinced that no substantial additional adjustments need to be done to the application of a mean variance framework to account for real estate specific risk and cost aspects. Based on our analysis, we are confident that the GPFG over time would benefit from increasing its allocation towards real estate to approximately at least 10 %. This could contribute to improving the risk return relationship of the portfolio, as measured through the Sharpe ratio. We have quantified the likely effect from an improvement in the risk reward ratio to 250 million NOK (45 mill USD) in additional return per year, with the current market capitalization of the fund.