This dissertation empirically and theoretically investigates three interrelated issues of market anomalies in interest rates derivatives and foreign exchange rates. The first essay models the spot exchange rate as a decomposition of permanent and transitory components. Unlike extant analysis, the transitory component could be stationary or explosive. The second essay examines the market efficiency hypothesis in the foreign exchange markets and relates the rejection of forward rate unbiasedness hypothesis to the existence of risk premium not to the failure of rational expectation. The third essay examines the behavior of short-term riskless rate and models the risk free rate as a nonlinear trend stationary process. While addressing these issues, these essays account for: (1) finite sample bias; (2) Unit root and other nonstationary behaviors; (3) the role of nonlinear trend; and (4) the interrelations between different behaviors. Several new results have been gleaned from our analysis; we find that: (1) the spot exchange rates display a very slow mean aversion behavior, which implies the failure of the purchasing power parity; (2) there are positive autocorrelations across the long horizons overlapping returns increases overtime and then begin to decline at a very long horizon period; (3) the short-term riskless rate displays a nonlinear trend stationary process which is closer to driftless random walk behavior; (4) modifying the mean reverting shortterm interest rates models to a nonlinear trend stationary shows an extreme improvement and outperforms all suggested models; (5) the traditional tests for rational expectations and market efficiency in the foreign exchange markets are subject to size distortions; (6) we relate the rejection of market efficiency in the foreign exchange markets documented across most currencies to the existence of risk premium not to the rejection of rational expectation hypothesis.

The topic of this Bachelor Thesis is “Which of these valuation methods provides the most accurate valuation forecast”. Assuming that the year is 2020, the goal of this thesis is to forecast the future stock prices of the fifty largest US-based companies on the Nasdaq stock exchange for 2021 and 2022. By using a quantitative method and looking ten years back at historical data. We determine which valuation method provides the most accurate stock price when conducted in a non-sector specific sample by comparing predicted prices to actual stock prices and discussing the results. There are several ways to evaluate a company and the ones being utilized in this thesis are the discounted cash flow valuation method, the price-to-earnings ratio method (equity multiple), and enterprise value to enterprise value before interest, tax, and depreciation (firm multiple). Our results show that when reviewing the valuations of multiple companies in different sectors the relative valuation methods provide better predictions with EV/EBITDA rather than the discounted cash flow method. This thesis provides the reader with a comprehensive overview of these different valuation methods and their effectiveness in providing valuation forecasts. The result of this thesis is beneficial for policymakers, investors, and financial analysts when forecasting future stock prices.