Refine
Has Fulltext
- yes (15)
Document Type
- Master's Thesis (13)
- Bachelor Thesis (2)
Keywords
- Benchmark Regulation (1)
- EURIBOR (1)
- Eigengeschäfte von Führungskräften (1)
- Ereignisstudie (1)
- Event Study (1)
- Event study (1)
- Financial Crises (1)
- Incentive Problem (1)
- LIBOR (1)
- LIBOR Scandal (1)
The aim of the paper is to implement the SRISK model on systemically relevant European financial institutions in order to investigate the applicableness as well as the usefulness regarding the supervisory of financial institutions. The model is implemented on 28 banks situated in Europe using the SRISK model proposed by Brownlees and Engle.
Half of the banks are considered as systemically relevant and the other half was chosen for reasons of comparison and further investigation. For the computation the CAPM Beta is used due to temporal reasons in the computation of the Dynamic Conditional Beta. It is shown that results defer but a ranking and monitoring is still possible.
The paper took a macro perspective to investigate the credit risk described by a number of major macroeconomics factors for the two major German retail banks (i.e. Deutsche Bank AG, Commerz Bank AG) in the past 10 years. The paper assessed the link between the macroeconomics and loan quality. The AR(1) macroeconomic credit risk models is applied to both banks in a aggregated and individual level. Both the sensitivity and the scenario analysis have been conducted. After shocking the macroeconomic factors in the model, the out –of -sample forecast suggests that the macro factors which have the most significant influence on the credit risk are household debt ratio, the import value and the DAX index price return.
This master thesis explores the spillovers and the dynamic conditional correlation between the Renewable energy and S&P 500 as well as Renewable energy and Oil & Gas indices for the period from October 2012 to December 2016. Its main purpose is to give an overview on the diversification potential of the Renewable energy.
A quantitative approach is adopted in the research. First, a cointegration analysis is implemented. The results from the Johansen’s test suggested that there are no co-integrated vectors in the long term between the Renewable energy and the other two markets, which implies that these markets do not share the same stochastic trend in the long run. At the same time, the linear dependencies obtained from the VAR(1) model suggest that in the short run there are evidence of spillovers from the equity to the Renewable energy market. In order to analyse conditional correlation in two high volatility periods – the Oil price shock in 2014 and the Chinese market turbulence in 2015, this master thesis utilizes a Dynamic Conditional Correlation model (DCC). The results imply an increase in the correlation between the Renewable energy and S&P 500 Equity and Renewable energy and the Oil & Gas index during the Chinese market turbulence and the Oil price shock.
The study has a number of implications for portfolio managers, policy makers and academic scholars.
Stock trading is a challenging decision-making problem that involves stock selection and asset management. Due to the complexity of stock market data, development of efficient models for trading that work well in the domain of out of sample data is very difficult.
Deep multilayer perceptron is trained to predict if a given company beats S&P 500 index in trading 30 days. The signal from the neural network is used for portfolio creation and trading strategy. The performance of the method is benchmarked against S&P 500, Markowitz
portfolio with Buy-and-Hold strategy as well as with rebalancing. Deep learning based portfolios outperform, as measured by accumulated return, Sharpe ratio. However, they underperform in the case of maximum drawdown, while still exhibiting better Calmar ratio than all the benchmarks .