dc.description.abstract | Today’s financial markets are currently experiencing stock index valuations close to all time high while low interest rates creates a negative outlook for fixed income-securities. Historically, stock markets will periodically experience downward corrections, and more and more market participants are starting to fear that the current five year bull-market is coming to an end soon. This thesis will therefore look into how an investor may position their portfolio to reduce the volatility without compromising the long-term return, both over a longer time-span and during times of increased uncertainty. An analysis is also done to try to predict which major, market-altering incidents that may occur over the coming years, while still recognizing that market-altering events are often characterized by being close to impossible to predict.
The historical cross-index relationship has been applied, and the index performance during times of market uncertainty was analyzed. This included an in-depth study of the financial crisis of 07-08 and the Eurozone crisis. The data was gathered using Reuters Datastream 5,1 with daily index observations from January 1st until the end of 2013. Combining this with a qualitative analysis of non-quantitative risk factors, as well as the likely future development, a new portfolio weighting was calculated, having the ability to achieve higher returns than the reference portfolio, while still experiencing lower historical volatility for the portfolio value. By using the GARCH(1, 1) equation, and calculating GARCH parameters for each relevant index, the forward development of volatility following a crisis could also be estimated.
The results indicated that institutional investors, such as Norway’s Sovereign Wealth Fund, should increase their allocation to riskier asset classes such as high yield-bonds, gold and equities of emerging markets, and still be able to reduce their overall risk exposure by allocating more than half of the portfolio to fixed income-securities. | nb_NO |