To extract usable information from financial data the prices of financial instruments must be summarized in an efficient manner. Typically price quotes are sampled at discrete and equidistant points in time to create a time series of prices at fixed times. However, alternative methods that instead utilize certain changes in the price data, such as price changes or drawdowns, could potentially create time series with more relevant information. This thesis builds upon previous research on so called ”directional changes” to establish scaling laws using such alternative sampling methods. This has been studied extensively for foreign exchange rates, and some of those results are replicated in this thesis. But here we also extend the results to a new domain of instruments, namely futures. In addition, data sampled with different methods is investigated for predictability using a simple classifier for forecasting trend direction. The main findings are that the aforementioned scaling laws hold for the time period investigated (2016-2020), and that using other methods than the typical discrete time method yields a more predictable time series when it comes to price trend. / För att utvinna användbar information ur finansiell data måste priser för finansiella instrument sammanställas på ett effektivt sätt. För kvoteringsdata görs detta vanligtvis genom att sampla priser vid diskreta tidsintervall för att får en serie av priser vid fixa tidsintervall. Alternativa metoder som använder händelser i prisdatan, som ändringar eller nedgångar, kan dock potentiellt skapa tidserier med mer relevant information. Den här uppsatsen bygger på tidigare forskning om s.k. ”directional changes” för att fastställa skalagar med sådana alternativa samplingmetoder. Det här görs inte bara för valutapar utan även för ett fåtal terminskontrakt. Utöver det undersöks prediktabiliteten hos data samplad med olika metoder med en enkel klassificerare för framtida trendriktning. De huvudsakliga resultaten är att de nämnda skallagarna håller för den undersökta tidsperioden (2016-2020) och att användandet av andra samplingsmetoder än diskret tidssampling resulterar i mer förutsägbara tidsserier när det kommer till pristrender.

[pt] Modelos de previsão cambial são frequentemente preteridos em relação a passeios aleatórios, porém o trabalho mostra que em certos casos, principalmente à médio e longo prazo, modelos simples de previsão cambial podem ser melhores do que passeio aleatório em países emergentes com câmbio livre. O trabalho também mostra que não há uma reversão do câmbio real à sua média de longo prazo e que seu movimento após um choque é feito quase todo pelo câmbio nominal. / [en] Foward looking exchange models are frequently deprecated when comparing to a random walk. This work notes that under certain cenarios, mostly at medium and long run, simple models can be more accurated than random walk for emerging countries with free floating exchange rates, though. This work also notes that there is no real exchange rate s mean-reverting at long run and that most of it s path, after a shock, is done via nominal exchange rate.

One of today’s financial trends is securitization. Evaluating Securitization risk requires some strong quantitative skills and a deep understanding of both credit and market risk. For international securitization programs it is mandatory to take into account the exchange-rates-related risks. We will see the di˙erent methods to evaluate extreme variations of the exchange rates using the Extreme Value Theory and Monte Carlo simulations. / Värdepapperisering är en av dagens finansiella trender. Att utvärdera vär-depapperisering risk kräver starka kvantitativa kunskaper och en förståelseför både kredit- och marknadsrisk. För internationell värdepapperisering ärdet obligatoriskt att hänsyn tas till valutarisker. Vi kommer att se de olika metoder för att utvärdera extrema variationer i valutakurser med hjälp av extremvärdesteori och Monte Carlo-simuleringar.

This thesis establishes methods to quantify and explain uncertainty through high-order moments in time series data, along with first principal-based improvements on the standard autoencoder and variational autoencoder. While the first-principal improvements on the standard variational autoencoder provide additional means of explainability, we ultimately look to non-variational methods for quantifying uncertainty under the autoencoder framework. We utilize Shannon's differential entropy to accomplish the task of uncertainty quantification in a general nonlinear and non-Gaussian setting. Together with previously established connections between autoencoders and principal component analysis, we motivate the focus on differential entropy as a proper abstraction of principal component analysis to this more general framework, where nonlinear and non-Gaussian characteristics in the data are permitted. Furthermore, we are able to establish explicit connections between high-order moments in the data to those in the latent space, which induce a natural latent space decomposition, and by extension, an explanation of the estimated uncertainty. The proposed methods are intended to be utilized in economic and financial factor models in state space form, building on recent developments in the application of neural networks to factor models with applications to financial and economic time series analysis. Finally, we demonstrate the efficacy of the proposed methods on high frequency hourly foreign exchange rates, macroeconomic signals, and synthetically generated autoregressive data sets. / Master of Science / This thesis establishes methods to quantify and explain uncertainty in time series data, along with improvements on some latent variable neural networks called autoencoders and variational autoencoders. Autoencoders and varitational autoencodes are called latent variable neural networks since they can estimate a representation of the data that has less dimension than the original data. These neural network architectures have a fundamental connection to a classical latent variable method called principal component analysis, which performs a similar task of dimension reduction but under more restrictive assumptions than autoencoders and variational autoencoders. In contrast to principal component analysis, a common ailment of neural networks is the lack of explainability, which accounts for the colloquial term black-box models. While the improvements on the standard autoencoders and variational autoencoders help with the problem of explainability, we ultimately look to alternative probabilistic methods for quantifying uncertainty. To accomplish this task, we focus on Shannon's differential entropy, which is entropy applied to continuous domains such as time series data. Entropy is intricately connected to the notion of uncertainty, since it depends on the amount of randomness in the data. Together with previously established connections between autoencoders and principal component analysis, we motivate the focus on differential entropy as a proper abstraction of principal component analysis to a general framework that does not require the restrictive assumptions of principal component analysis. Furthermore, we are able to establish explicit connections between high-order moments in the data to the estimated latent variables (i.e., the reduced dimension representation of the data). Estimating high-order moments allows for a more accurate estimation of the true distribution of the data. By connecting the estimated high-order moments in the data to the latent variables, we obtain a natural decomposition of the uncertainty surrounding the latent variables, which allows for increased explainability of the proposed autoencoder. The methods introduced in this thesis are intended to be utilized in a class of economic and financial models called factor models, which are frequently used in policy and investment analysis. A factor model is another type of latent variable model, which in addition to estimating a reduced dimension representation of the data, provides a means to forecast future observations. Finally, we demonstrate the efficacy of the proposed methods on high frequency hourly foreign exchange rates, macroeconomic signals, and synthetically generated autoregressive data sets. The results support the superiority of the entropy-based autoencoder to the standard variational autoencoder both in capability and computational expense.

Bakgrund: Aktiemarknaden är viktig för att företag ska kunna samla kapital och möjligheten för investerare att få avkastning. Utländskt ägande är betydande, vilket ökar likviditeten men samtidigt sårbarheten. Valutakurser påverkar aktiemarknaden, särskilt för exportföretag. En svag krona gör svenska investeringar mer attraktiva och gynnar exporten. Sambandet mellan valutakurser, aktiemarknaden och utländskt ägande är viktigt för investerare. Syfte: Studiens syfte är att undersöka det dynamiska sambandet mellan valutakurser, utländskt ägande och den svenska aktiemarknaden. Dessutom avser den att beskriva om det går att förutse marknadsrörelser med hjälp av valutakurserna. Metod: Studien använder sig av kvantitativ metod och använder sig av Pearson korrelationsanalys samt multipel regression. Teori: Studien utgår från effektiva marknadshypotesen och behavioral finance.  Slutsatser: Det visade sig att 4 av 5 valutor har en signifikant korrelation till OMXSPI, där den norska kronan är den enda som saknar samband. Korrelationen är dock för svag för att kunna säga att den kan hjälpa till att förutsäga rörelser i aktiekursen. Samtliga valutor har ingen korrelation med det utländska ägandet. Därför går det inte heller att med hjälp av valutorna förutsäga förändringar i det utländska ägandet på marknaden. / Background:The stock market is important for companies to raise capital and an opportunity for investors to get returns. Foreign ownership is significant, which increases liquidity but at the same time volatility. Exchange rates affect the stock market, especially for export companies. A weak krona makes Swedish investments more attractive and favors exports. The relationship between exchange rates, the stock market and foreign ownership is important to investors. Purpose: The purpose of the study is to investigate the dynamic relationship between exchange rates, foreign ownership and the Swedish stock market. In addition, it intends to see if it is possible to predict market movements using the exchange rates. Methodology: In this study a quantitative method is used in the form of Pearson's correlation analysis and multiple regression analysis.  Theory: The study is based on the efficient market hypothesis and behavioral finance. Conclusions: It turned out that 4 out of 5 currencies have a significant correlation to OMXSPI, with the Norwegian krone being the only one without correlation. However, the correlation is too weak to say that it can help predict movements in the share price. All currencies have no correlation with foreign ownership. Therefore, it is also not possible to use the currencies to predict changes in foreign ownership in the market.

This study examined the effect of real exchange rate misalignments on economic growth in sub-Saharan Africa (SSA) by employing 15 countries. The sample is subdivided into 7 lowincome countries and 8 middle-income countries. The dataset spans 41 years covering the period 1970-2010. The study examined this broad issue in piecewise fashion. In the first part, the study examined the validity of the Balassa-Samuelson Hypothesis (BSH) using a simplified regression model and within-effects estimations. The study found a negative and highly significant coefficient of the relative productivity term for the two subsamples (i.e. low-income SSA countries and middle-income SSA countries), in addition to the full sample. Thus, the study found a well-established BSH for the SSA countries considered. Second, the study examined the impact of the real exchange rate undervaluation on economic growth using a standard regression model with key control variables. The study constructed an index of undervaluation, following Rodrik (2008). The study also constructed a Hodrick- Prescott based undervaluation index in order to evaluate the robustness of the main undervaluation index. Generally, the study found undervaluation to promote growth and overvaluation to reduce it. The study found the effect of undervaluation on economic growth to weaken as countries migrate from the low-income bracket to the middle-income bracket. Moreover, the study examined whether the choice of the undervaluation measure mattered. The study found the choice of the undervaluation measure to matter. The Rodrik-type index appeared to overestimate the size of the impact of undervaluation on economic growth. Finally, the study examined whether the impact of undervaluation on growth was linear. The evidence showed that the impact of undervaluation on growth was linear, at least, for this study. The linear impact of real exchange rate movements on economic growth implied that undervaluation enhanced economic growth just as overvaluation hindered it. / Economics / D. Phil. (Economics)

by Liu Shiang Ling. / Thesis (M.B.A.)--Chinese University of Hong Kong, 1997. / Includes bibliographical references (leaves 48). / ABSTRACT --- p.ii / TABLE OF CONTENTS --- p.iii / LIST OF TABLES --- p.v / CHAPTER / Chapter I. --- INTRODUCTION --- p.1 / Chapter II. --- CHINA'S RENMINBI SYSTEM --- p.3 / Overview and Historical Development --- p.3 / 1953-1972 --- p.3 / 1973-1978 --- p.4 / 1979-1993 --- p.4 / 1994-1996 --- p.6 / The Exchange Rate Determination --- p.7 / Demand & Supply --- p.7 / The Balance of Payments --- p.8 / Inflation --- p.10 / Interest Rate --- p.10 / The Condition of Economic Development --- p.11 / Expectation --- p.11 / The Fiscal Policy --- p.12 / Forecasting Exchange Rate --- p.12 / Fundamental Analysis --- p.14 / Chapter III. --- CHINA'S PROPERTY MARKET --- p.15 / The History of China's Property Market --- p.15 / Pre-1949 --- p.15 / 1949-1979 --- p.15 / 1979-1991 --- p.16 / 1992-1996 --- p.18 / China's Real Estate Market Overview --- p.20 / Shanghai's Property Market --- p.21 / Risk Analysis --- p.22 / Return --- p.22 / Risks --- p.23 / Chapter IV. --- THEORETICAL ANALYSIS --- p.25 / Currency Conversion and Exchange Rate Risks --- p.25 / Hypothesis --- p.26 / The Short Run --- p.26 / The Long Run --- p.27 / Chapter V. --- RESEARCH METHODOLOGY AND FINDINGS --- p.29 / Targets of Study --- p.29 / Questionnaire --- p.29 / Chapter VI. --- ANALYSIS AND CONCLUSIONS --- p.30 / Primary Source Data --- p.30 / Secondary Source Data --- p.32 / Analysis --- p.34 / The Short Run --- p.34 / The Long Run --- p.35 / Political Conditions --- p.35 / Economic Conditions --- p.36 / Social Conditions --- p.37 / Market Conditions --- p.37 / Land Investment --- p.38 / Conclusion --- p.39 / Chapter VII. --- RECOMMENDATIONS AND ADVICE TO DEVELOPERS --- p.40 / The Depreciation of Renminbi --- p.40 / The Appreciation of Renminbi --- p.41 / The Chinese Government Policy --- p.42 / Long-Term Consideration --- p.43 / APPENDIX --- p.45

Chen Chen. / Thesis (M.Phil.)--Chinese University of Hong Kong, 2001. / Includes bibliographical references (leaves 122-129). / Abstracts in English and Chinese. / ABSTRACT --- p.ii / ACKNOWLEDGEMENT --- p.iv / TABLE OF CONTENTS --- p.v / LIST OF FIGURES --- p.vii / LIST OF TABLES --- p.viii / Chapter / Chapter I. --- INTRODUCTION --- p.1 / Chapter 1.1 --- Introduction --- p.1 / Chapter 1.2 --- Objectives and Motivation --- p.5 / Chapter 1.3 --- The Asian Crisis --- p.9 / Chapter 1.4 --- Procedures and Findings --- p.18 / Chapter 1.5 --- Summary --- p.20 / Chapter II. --- LITERATURE REVIEW --- p.21 / Chapter 2.1 --- Definition and Determinants --- p.21 / Chapter 2.2 --- Measurement Model --- p.25 / Chapter 2.3 --- Exchange Rate Fluctuation and Market Value of the Firm --- p.28 / Chapter 2.3.1 --- Exchange Rate Fluctuation and Stock Return --- p.28 / Chapter 2.3.2 --- Some Problems of the Measurement Model --- p.31 / Chapter 2.4 --- Exchange Rate Fluctuation and Market Risk of the Firm --- p.42 / Chapter 2.5 --- Summary --- p.45 / Chapter III. --- HYPOTHESES，METHODOLOGY & DATA --- p.47 / Chapter 3.1 --- Hypotheses --- p.47 / Chapter 3.2 --- Research Design --- p.50 / Chapter 3.3 --- Sample Selection --- p.56 / Chapter 3.3.1 --- Selection of Sample Group --- p.56 / Chapter 3.3.2 --- Selection of Control Group --- p.61 / Chapter 3.3.3 --- Comparison of Two Groups --- p.62 / Chapter 3.4 --- Data and the Measurement of the Variables --- p.64 / Chapter 3.5 --- Summary --- p.67 / Chapter IV. --- EMPIRICAL RESULTS AND DISCUSSION --- p.68 / Chapter 4.1 --- Exchange Rate Variability and Stock Return Volatility --- p.68 / Chapter 4.2 --- Exchange Rate Variability and Market Risk --- p.81 / Chapter 4.3 --- Interpretations --- p.87 / Chapter 4.3.1 --- Phenomenon 1: Cost of Equity and Net Cash Flows --- p.89 / Chapter 4.3.2 --- Phenomenon 2: Increased Return Variability and the US Stock Market Return --- p.92 / Chapter 4.4 --- Alternative Explanation --- p.96 / Chapter 4.5 --- Summary --- p.99 / Chapter V. --- CONCLUDING REMARKS --- p.100 / APPENDICES / APPENDIX 1. Firm Lists --- p.105 / APPENDIX 2. Estimates of CAPM Betas --- p.115 / BIBLIOGRAPHY --- p.122

自從2007年發生金融海嘯後，歐盟貨幣整合問題持續被討論，其中，歐盟成員國各國經濟狀況差距過大被認為是這次經濟危機爆發出種種問題的主要原因。而近幾年，在國際其他地區區域性整合刺激之下，亞洲各國開始重視簽屬自由貿易協定的重要性，甚至未來朝區域貨幣整合作發展。 此論文主要是探討東亞國家是否具備成立亞元區的條件。我們探討東亞各國2004-2015年間對美元匯率以及消費者物價指數(CPI)變化是否有一致的趨勢，主要使用MST及HC、DCG tree機器學習分群演算法作為分析工具，對於東亞十個國家，包含台灣、日本、中國、南韓、香港、印尼、泰國、馬來西亞、菲律賓、新加坡以及用於對照的世界上其他國家作分析，以2008年金融海嘯以及2013年日本提出新經濟政策作為分界點，分為三個時期做討論。 研究結果顯示，三個時期，東亞國家匯率變動相較於世界其他國家皆較有一致的趨勢，而更進一步檢視，我們發現在日本提出新經濟政策前後，東亞國家匯率結構發生明顯的變化。物價指數方面同樣也是東亞國家有較集中的趨勢，但物價指數不向匯率，不同時期並無明顯變化。我們進而推論東亞國家以符合成立亞元區最基本的條件，匯率和物價指數的一致性，但是有幾個結果不能忽略，日本對於東亞國家匯率結構的影響力以及人民幣、日圓等強勢貨幣匯率變動趨勢在分析結果中和其餘東協國家還是有些差距，這些可能是在未來貨幣整合過程中市所需要注意的部分。 / After the global financial crisis, European Union (EU) faced a range of economic and political problems, including the Greek debt crisis and the upcoming UK referendum on EU membership. People start to think what factors cause these problems. A large literature has emerged discussing this issue and examining the future directions of monetary and exchange rate agreements of many countries in the region. One of the popular questions has raised researchers’ interests: Are East Asian countries possible to be the next monetary integrated area? This paper investigates the similarity of exchange rates and consumer price index (CPI) in the East Asian region, Taiwan, Japan, China, Hong Kong, South Korea and five ASEAN member countries by cluster analysis. We use HC tree and DCG tree to discuss the structures of these two macro economic factors and whether major economic events between 2004~2015 such as global financial crisis and Japan’s new economic policy influenced the structure. The exchange rates and CPI are both found to be more consistent among East Asian countries, including that these countries already have the basic condition to be the monetary integrated area. However, it is important to pay attention to Japan’s new economic policy, which influenced the structure a lot, and that Japanese yen and Chinese yuan still have gaps with ASEAN member countries.

This collection of essays examines the topic of macroeconomic stabilisation in an international context, focusing on monetary policy, capital controls and exchange rates. Chapter 1, written in collaboration with Giancarlo Corsetti and Joao Duarte, reconsiders the effects of common monetary policy shocks across countries in the euro area, using a data-rich factor model and identifying shocks with high-frequency surprises around policy announcements. We show that the degree of heterogeneity in the response to shocks, while being low in financial variables and output, is significant in consumption, consumer prices and macro variables related to the labour and housing markets. Mirroring country-specific institutional and market differences, we find that home ownership rates are significantly correlated with the strength of the housing channel in monetary policy transmission. We document a high dispersion in the response to shocks of house prices and rents and show that, similar to responses in the US, these variables tend to move in different directions. In Chapter 2, I build a two-country, two-good model to examine the welfare effects of capital controls, finding that under certain circumstances, a shut-down in asset trade can be a Pareto improvement. Further, I examine the robustness of the result to parameter changes, explore a wider set of policy instruments and confront computational issues in this class of international macroeconomic models. I document that within an empirically relevant parameter span for the trade elasticity, the gains from capital controls might be significantly larger than suggested by previous contributions. Moreover, I establish that a refined form of capital controls in the shape of taxes and tariffs cannot improve upon the outcome under financial autarky. Finally, results show that the conjunction of pruning methods and endogenous discount factors can remove explosive behaviour from this class of models and restore equilibrating properties. In Chapter 3, I use a panel of 20 emerging market currencies to assess whether a model that combines fundamental and non-fundamental exchange rate forecasting approaches can successfully predict risk premia (i.e. currency excess returns) over the short horizon. In doing so, I aim to overcome three main shortcomings of earlier research: i) Sensitivity to the chosen sample period; ii) seemingly arbitrary selection of explanatory variables that differs from currency to currency; and iii) difficulty in interpreting forecasts beyond the numerical signal. Based on a theoretical model of currency risk premia, I use real exchange rate strength combined with indicators for carry, momentum and economic sentiment to homogeneously forecast risk premia across all 20 currencies in the sample at a monthly frequency. In doing so, the model remains largely agnostic about structural choices, keeping arbitrarily imposed restrictions to a minimum. Results from portfolio construction suggest that returns are significant and robust both across currencies as well as over time, with Sharpe Ratios in out-of-sample tests above 0.7.