Optimal Reserve Holdings, Strategic Asset Allocation and Multiple-Goal Investment Plan for Sovereign Wealth Fund of China

Abstract

This thesis expounds China’s foreign reserve policy and the investment management of the reserves in a behavioural approach. The research provides a behavioural explanation of China’s reserve accumulation, which is based on the optimal decision making under uncertainty. Then the thesis proposes a multiple-goal framework for strategic asset allocation of China’s reserve management and for the investment decision of Chinese Sovereign Wealth Fund (SWF).
The research first tackles the reserve accumulation puzzle in China, by incorporating loss aversion and narrow framing into the utility maximisation of the representative agent who makes the decision of wealth allocation between consumption and saving under uncertainty. Due to China’s policy maker’s subscription to promoting GDP growth as the primary political goal, it is reasonable to assume that the policy maker as a representative agent derives utility not only from consumption but also from fluctuations of the value of GDP/income. This agent evaluates the possible uninsured risk of GDP fluctuation narrowly and tends to exhibit the attribute of loss aversion relative to her growth expectation as the reference point. Under the influence of loss aversion and narrow framing, the more the policy maker cares about GDP growth, the more she needs reserve assets as a precautionary means that may provide self-insurance against uninsured income risk. Such cognitive biases enhance the agent’s precautionary motive for foreign reserves in an uncertain world, which in turn leads her to believing in an optimal level of foreign reserves that is higher than that under conventional models with rational agents. Hence, this heightens the accumulation of foreign reserves in China.
Second, this thesis develops a new construction of strategic asset allocation for central banks’ management of foreign reserves by way of embedding the Black-Litterman (B-L) model into the mean variance mental accounting (MVMA) framework. While the MVMA measure suggests a multiple-objective framework that may embrace the traditional objectives of reserve management, i.e. safety, liquidity and profitability, it is based on the mean-variance approach, which suffers from profound deficiencies such as the unrealistic objective function that it relies on and the tendency that the methods are prone to undue influences of outliers. So, the B-L model is applied in this study to form forward-looking return forecasts. This method allows us to overcome the error-maximising influences of the mean-variance optimization. Furthermore, one can combine the implied equilibrium excess returns as investors’ investment views to form priors for Bayesian estimation. The optimal asset allocation then can be derived in this framework, which is applied to practical use in the context of China.
The third main Chapter of this thesis concerns the investment of China’s sovereign wealth fund (SWF). The establishment of the Chinese SWF can be regarded as an optimal policy response to the changing economic conditions facing China. This fund as a special investment vehicle proves very useful for China to focus on the returns objective of managing the reserve assets, on top of the safety and liquidity objectives. This is especially important in a low yield international environment. To help achieve the yield objective, this Chapter develops further the behavioural portfolio model cum the Black-Litterman method to derive the optimal asset allocation for China’s sovereign wealth fund.