In a nutshell
Allocation of SWF investments should take account of the rates of return on different assets classes, oil price dynamics and how the size of the fund evolves with respect to the national economy.
The current allocation of 60% in stocks and 40% in bonds in the largest MENA SWFs is optimal given moderate levels of risk aversion and the current volatility of oil prices.
The optimal rate of consumption out of the fund decreases with both risk aversion and oil price volatility.
The sovereign wealth fund (SWF) model, and its proliferation in the last two decades, has been a marked development in the way that national savings can be harnessed to achieve multiple objectives. Serving primarily as vehicles for precautionary saving and intergenerational wealth transfer, they have become significant players in international financial markets, with about US$7.9 trillion in assets under management.
Nearly two thirds of existing SWFs are based on the proceeds from oil exports. Indeed, the phenomenal growth in their size over the last decade is largely due to the massive current account surpluses enjoyed by oil exporters during the extended oil price boom over the period 2007-2014, save for the price drop of 2009.
The biggest SWF – the Government Pension Fund (Global) of Norway – has more than US$1 trillion in assets, and it owns roughly 1.3% of the stocks of every major company around the globe. Following Norway, Kuwait, Qatar, Saudi Arabia and the United Arab Emirates own the largest oil-based SWFs.
For these economies, the SWF assets under management are either at par or exceed the value of their GDP. This highlights their potential role in macroeconomic stabilisation. The drawdowns witnessed in some of the funds, particularly in Saudi Arabia, in the face of the sharp decline in oil prices in 2014 is a case in point.
The focus of this column is how oil-based SWFs should allocate their investment across different asset classes, such as stocks, bonds and real estate. While, at first glance, this seems like the standard asset allocation problem that has been widely studied in the finance literature since the pioneering work of Merton (1969, 1971), the workings of SWFs are different.
The primary reason for the difference is that growth in SWF assets over time comes from two distinct sources:
- The return on a fund’s investment.
- The injection of new capital into the fund from the proceeds of oil exports.
Our work in Moutanabbir and Noureldin (2018) focuses on exploiting the interdependence between these two sources of growth since different asset classes, especially stocks, tend to be correlated with oil prices. For the SWF manager, this offers a window of opportunity to implement a dynamic asset allocation strategy, which allows for hedging against adverse movements in oil prices.
So, how are oil-based SWFs currently allocating their investments? There seems to be a convergence towards a 60-40 allocation, with 60% allocated to stocks and 40% allocated to bonds and alternative investments, including real estate and direct equity investments in infrastructure projects.
More recently, the Norwegian government approved a gradual move to a 70-30 allocation, thereby seeking to invest a larger share in stocks. This shows that the primary objective of these funds is to maximise risk-adjusted return, despite the existence of some evidence that they have also been historically used to fulfil strategic objectives related to the development of know-how in particular industries (Dyck and Morse, 2011; Bernstein et al, 2013).
With a large share of their assets invested in stocks, SWFs are indeed taking on financial risk and the return on a fund’s investment is likely to be more volatile. But given their long investment horizon, this is exactly the optimal thing to do.
In a way, this resembles the optimal investment strategy of a young employee saving for retirement. Given that she will be retiring in 40 years’ time, her investment plan can withstand boom and bust cycles in stock markets. As she nears retirement, it is optimal to reduce her exposure to stocks and allocate more to bonds as a low-volatility alternative.
But for a SWF, this age transition does not apply, hence it is optimal to remain heavily invested in stocks to achieve long-term growth in the fund. Some of these insights are discussed in Campbell et al (2003) and Viceira (2001), on which our analysis builds.
So, how should a fund allocate their assets in a manner that also allows for hedging against adverse oil price shocks? The answer to this question varies with the following key parameters of our analysis:
- The degree of dependence between oil price movements and that of the other asset classes in which the SWF is invested.
- The extent of oil price volatility.
- The level of maturity of the fund.
We consider a fund to be mature when its value has become so large such that the annual stream of oil income that is added to the fund is negligible in comparison. The SWF of Norway is perhaps the best example of a mature fund since annual oil revenue represents a mere 2% of its value, compared with about 20% in the case of Saudi Arabia.
Our derived optimal asset allocation dynamic rule shows that the current 60-40 allocation observed in the largest MENA SWFs is indeed an optimal allocation, given moderate levels of risk aversion and the current level of volatility in oil prices. Thus, these funds seem to be behaving optimally in a manner that maximises risk-adjusted returns over a long-time horizon.
But given that oil price volatility has increased in recent months, it is optimal for the funds to increase the allocation to stocks as a hedging mechanism given the historical negative correlation between stocks and oil prices. The Norwegian fund is moving in this direction.
We also specify a rule for the optimal rate of consumption out of the fund – that is, how much the government can take out of the fund to support fiscal operations without compromising future growth in the fund’s value. This rule is also dynamic and intertemporal since high current consumption will lower the fund’s assets and thus necessarily lower the consumption of future generations.
Our derived rule indicates that the optimal rate of consumption out of the fund decreases with both risk aversion and oil price volatility. Indeed, we find the spending rule in Norway (historically 4% since 2001, and now reduced to 3%) to be optimal given an increase in the fund’s maturity and a general increase in oil price volatility since 2014.
The key policy message of our analysis is that the allocation of SWF investments should be viewed dynamically within a responsive framework that takes account of the rates of return on different assets classes, oil price dynamics and how the size of the fund evolves with respect to the national economy.
Bernstein, S, J Lerner and A Schoar (2013) ‘The Investment Strategies of Sovereign Wealth Funds’, Journal of Economic Perspectives 27: 219-38.
Campbell, JY, YL Chan and LM Viceira (2003) ‘A Multivariate Model of Strategic Asset Allocation’, Journal of Financial Economics 67: 41-80.
Dyck, A, and A Morse (2011) ‘Sovereign Wealth Fund Portfolios’, Chicago Booth Research Paper No. 11-15.
Merton, RC (1969) ‘Lifetime Portfolio Selection under Uncertainty: The Continuous Time Case’, Review of Economics and Statistics 51: 247-57.
Merton, RC (1971) ‘Optimum Consumption and Portfolio Rules in a Continuous-time Model’, Journal of Economic Theory 3: 373-413.
Moutanabbir, K, and D Noureldin (2018) ‘Optimal Asset Allocation and Consumption Rules for Commodity-Based Sovereign Wealth Funds’, ERF Working Paper No. 1172.
Viceira, LM (2001) ‘Optimal Portfolio Choice for Long-horizon Investors with Nontradable Labor Income’, Journal of Finance 56: 433-70.