Essays on risk-free interest rate and Beta

The risk-free interest rate is a commonly utilized parameter in modern finance. A review of existing studies indicates that various approaches are adopted when selecting the risk-free rate. Factors contributing to these differences include the investor's perception of risk, the modeling period, and prevailing market conditions. It is crucial to acknowledge the variability in selecting the risk-free rate as a reference point, as it can significantly influence investment decisions. However, research focusing specifically on this topic remains relatively scarce. The first essay of this thesis aims to highlight the potential effects of differing preferences for the risk-free rate among financial market participants, employing methodologies based on both intermittent and expanding periods. The findings suggest that variations in preferences for the risk-free rate can result in differences in optimal portfolio allocations. One of the most critical parameters guiding practitioners in modern finance is Beta. Since it is an unobservable characteristic, predicting Beta requires careful estimation. The fundamental approach to estimating Beta involves predicting covariances and variances based on historical return data. However, a significant challenge in calculating the Beta coefficient is its variation over time. It is essential to find a method that accounts for this variability in Beta estimation. In doing so, identifying whether Beta estimates are influenced by the chosen prediction window and the frequency of the data set is essential, as it offers a new perspective in the relevant research area. The second essay of this thesis examines the performance of Rolling and Recursive Regression methods in predicting Beta over daily and weekly frequencies within the windows of 252, 126 trading days, and 52, 26 trading weeks. The results indicate that the Rolling regression method performs more successfully, that daily data frequency enables more consistent Beta predictions compared to weekly data, and that the 126-day forecast window provides the most optimal Beta estimate. When forecasting systematic risk, examining the factors influencing Beta is a valuable step in understanding systematic risk. This approach directly contributes to financial decision-makers and practitioners behaving in a manner that aligns with economic and financial realities by analyzing the direction and magnitude of the impact of firm/sector-specific factors on Beta. Identifying firm-specific factors affecting Beta allows for the detection of risk characteristics unique to a firm and sector, offering valuable information to both investors and company managers when making decisions. In the third essay of this thesis, various firm-specific variables are used to analyze their impact on Beta through panel data analysis. The findings indicate that the factors influencing Beta vary across different sectors.