The competitive electricity market faces low margins while the energy transition entails volatile electricity prices and major risks in the market. Along with these problems, there are challenges in optimizing energy portfolios, as they are based on maximized return and minimized risks. Portfolio managers handle electricity contracts for customers with the aim of offering competitive electricity contracts and at the same time achieving profit. This study therefore aims to investigate how electricity trading companies can optimize the energy portfolio for corporate customers in the Swedish electricity market. In addition, it is analyzed how these companies can adapt portfolio optimization to the challenges of the electricity market and what kind of digital systems are needed to support portfolio optimization. Previous studies in the subject have mainly focused on a quantitative approach. In this study on the other hand, a qualitative method is used. This in order to use interviews, models and theories to investigate how electricity trading companies can optimize the portfolio for their corporate customers. The study is divided into an empirical part that includes interviews with 12 electricity trading companies, as well as a theoretical part. The theoretical part deals with theories for risk management, including portfolio optimization, price hedging with electricity derivatives and the Risk Management Payoff model. In addition, the Customer Relationship Management (CRM) model is included, with the aim of strengthening relationships. The theoretical framework has been used to analyze the empirical data of the study and has resulted in conclusions that answer the research questions. The results show that in order to achieve energy portfolio optimization, futures contract is the derivative that provides the lowest risk and can be used as the optimal hedging tool. For larger corporate customers, however, forward contract is a more suitable hedging tool. This is because larger customers prefer to secure monthly contracts, rather than daily settled contracts with futures. To support portfolio management with electricity contracts, there is a need for support systems. Electricity trading companies are in need of a common system for both financial and physical trading to enable easier management. In addition, security functions in systems as well as more informative and customer-friendly systems are of interest. Portfolio optimization is customer centric, thus a strong relationship between portfolio manager and customer, with good trust and reputation, is important. Finally, adaptation in portfolio management with diversification and automation with AI, can be a way to achieve competitiveness in the electricity market also in the future.

The thesis deals with the optimization of the selected investment portfolio. Solver suggests automated investment model that will use advanced algorithms based on artificial intelligence and principles of technical analysis. Optimization of parameters and verifying the performance of the investment model is realized on historical market data. The result of this thesis is optimized investment model with an emphasis on maximizing profits and stability. The thesis is realized in an environment Python programming language and freely available analytical libraries.

Diploma thesis focuses on the issue of an appropriate selection of securities and the subsequent establishment of a portfolio of these securities. Follow detailed discussion about analysis of portfolio and investor’s preferences. Below is a description of the CAPM model, its assumptions and usage of this model to build a portfolio. Then there is the actual calculation of characteristics of securities traded on the Prague Stock Exchange and on the basis of these calculations is made the proposal of several portfolios and their evaluation.

This master‘s thesis is focused on creating an application, which would suggest an optimal portfolio of shares from SPAD Stock Market Praha. The application is based on the CAPM model, which is also described in this paper. There is a calculation of securities characteristics and specific portfolio diversification is suggested. The application also allows a user to simulate investments based on his requirements.

The performance of an optimal-weighted portfolio strategy is evaluated when transaction costs are penalized compared to an equal-weighted portfolio strategy. The optimal allocation weights are found by maximizing a modified Sharpe ratio measure each trading day, where modified refers to the expected return of an asset in this context. The leverage of the investment is determined by a conditional expectation estimate of the number of portfolio assets of the next-coming day. A moving window is used to historically measure the transition probabilities of moving from one state to another within this stochastic count process and this is used as an input to the estimator. It is found that the most accurate estimate is the actual trading day’s number of portfolio assets and this is obtained when the size of the moving window is one. Increasing the penalty parameter on transaction costs of selling and buying assets between trading days lowers the aggregated transaction cost and increases the performance of the optimal-weighted portfolio considerably. The best portfolio performance is obtained when at least 50% of the capital is invested equally among the assets when maximizing the modified Sharpe ratio. The optimal-weighted and equal-weighted portfolios are constructed on a daily basis, where the allowed VaR0:05 is €300 000 for each portfolio. This sets the limit on the amount of capital allowed to be invested each trading day, and is determined by empirical VaR0:05 simulations of these two portfolios.

In this thesis, an optimal investment problem is studied for an investor who can only invest in a financial market modelled by an Itô-Lévy process; with one risk free (bond) and one risky (stock) investment possibility. We present the dynamic programming method and the associated Hamilton-Jacobi-Bellman (HJB) equation to explicitly solve this problem. It is shown that with purification and simplification to the standard jump diffusion process, closed form solutions for the optimal investment strategy and for the value function are attainable. It is also shown that, an explicit solution can be obtained via a finite training of a neural network using Stochastic gradient descent (SGD) for a specific case.

The aim of this thesis is to explore the role of bitcoin in an investment portfolio. The paper examines the nature of bitcoin and additionally how bitcoin compares to gold when included in an investment portfolio. This report uses the historical value of bitcoin and investigates with a Mean-Variance model how the risk-adjusted return of an optimized portfolio is affected when bitcoin is a constituent. By comparing Sharpe Ratios from the optimized portfolios, a conclusion can be drawn as to whether bitcoin affects the maximum Sharpe ratio or the global minimum variance point. Our study suggests that including bitcoin in an investment portfolio increases the risk-adjusted return of the portfolio. In addition, portfolios optimized with bitcoin outperform the market. Further, we conclude that bitcoin has a relatively high correlation as compared to gold with the assets in the study. Hence, bitcoin is not the new gold.

This thesis analyzes the mean variance optimization problem with respect to cardinalityconstraints. The aim of this thesis is to figure out how much of an impact transactionchanges has on the profit and risk of a portfolio. We solve the problem by implementingmixed integer programming (MIP) and solving the problem by using the Gurobi solver.In doing this, we create a mathematical model that enforces the amount of transactionchanges from the initial portfolio. Our results is later showed in an Efficient Frontier,to see how the profit and risk are changing depending on the transaction changes.Overall, this thesis demonstrates that the application of MIP is an effective approachto solve the mean variance optimization problem and can lead to improved investmentoutcomes.

Hoarding a too large cash reserve is often unfavourable due to lost investment opportunities. Similarly, an insufficient cash reserve can be detrimental, as one might fail to meet payment obligations. Finding the optimal balance is nothing that is done in the blink of an eye, particularly when the underlying variable is stochastic, e.g., the life span of a human being. Resscapital is a fund manager investing in the secondary and tertiary markets for life insurance policies, also known as the life settlements market. They are currently on a mission to set up a closed-end fund where one of the main challenges is balancing the invested capital and the amount of capital set aside for payment obligations. The stochastic nature of life insurance policies entails the difficulty to foresee future premium payments and face value payouts. Without a model forecasting the cash flows, decisions regarding the cash reserve are based on nothing better than a guesstimate. Thus, with the aim to help determine the minimum cash reserve required to cover the payment obligations, this thesis was initiated. By developing a methodology based on general theory, the objective of this thesis is reached and the purpose fulfilled. The proposed model uses Monte Carlo simulation to generate scenarios that eventually creates a distribution of required cash reserves. Following the inversion principle, the remaining lifetime for each and every individual is simulated from their empirical distribution of survival probabilities, respectively. After simulating the occurrences of demise, an algorithm builds up the cash flows for the entire fund term for that specific scenario based on predetermined parameters. Since cash flows stem from both assets and management, the portfolio must be revalued continuously, demanding a gradual evaluation of the cash flows during the fund term. Repeated a large number of times, the quantile corresponding to any confidence level is attained by using a Value at Risk methodology. Analysis of the results and sensitivity analysis on the parameters provides a deeper understanding of the underlying factors, revealing, among other things, that longevity risk for policies with short life expectancy is a key driver of the required cash reserve. Furthermore, validation of the model shows that the results are sufficient and serve the purpose well.

ESG investing has been a hot topic during several years and there have been numerousstudies examining the relationship between sustainable assets and non-sustainable assetsincluding green bonds, social bonds, environmental bonds, ESG-bonds and ESG indices;conventional bonds, S&P 500, common stocks and non-ESG indices. During negative marketshocks several ESG stocks and indices have been shown to outperform common stocks andindices. Green bonds demonstrated an asymmetric relationship to other assets providinginvestors with an opportunity for diversification. We’ve looked at the relationship andperformance of sustainable assets and non-sustainable assets by using Markowitz portfoliometrics and Engle Rs’ DCC-GARCH. Our findings propose green bonds and treasuries toprovide hedging and diversification opportunities during crises but demonstrate sustainablefixed income assets to underperform non-sustainable fixed income assets during the COVID19 market shock as opposed to previous studies.