根據研究指出 ETF 資產規模近幾年快速成長，元大台灣卓越 50 基金因市場 規模大等優勢受到投資人的青睞，賴以巨量資料的發展使得文字探勘技術成熟， 故本研究希冀提出一套情感分析的價格預測模型，提升投資者的報酬率。 過往學者以文章中的單詞作為文字探勘的分析單位，常會產生同義詞、多義 詞的問題，因此提出情感型態分析的監督式學習方法建立模型。另外為了解決監 督式學習難以取得訓練資料的限制，本研究混合非監督式學習方法進行主題分群 與情緒傾向標注。 本研究建立台灣股市新聞文本資料集，並篩選熱門議題詞詞庫，進行非監督 式的 LDA 主題模型，發現在 2016 年總統選舉期間，媒體對於公司相關議題的注 意力降低，使得相關的文本數量大幅減少;另外在情緒傾向標注階段，因混和了 NTUSD、知網及自行擴充演算法的情感詞庫，能夠將 10%中性詞彙產生極性判 斷、96%的文本標注情緒傾向。 視覺化工具分析結果指出，DIF-MACD 能夠預測台灣卓越 50 基金的長期走 勢，而新聞情緒指數則在短期的價格波動上表現良好，且在主題模型分群中，總 體經濟、公司維運類別的新聞情緒指數具有約 1-2 日領先指標特性，對於後續的 價格預測模型有所助益。 在監督式情感分析方法，為解決上述同義詞、多義詞的問題，本研究採用型 態分類模型於中文文本，並與向量空間模型、支援向量機等方法做比較。實驗結 果指出優化的型態分類模型，並結合台灣加權股價指數，表現相對良好，F1- Measure 可達 85%。進一步討論新聞情緒對於價格預測的重要性，發現在非交易 時間序列中的新聞情緒，能夠對 0050 的價格波動產生影響。 / The past research points out that the scale of ETF assets has been growing rapidly in recent years. Yuanta Taiwan Top 50 ETF is popular with investors because of the advantages of large market scale. Through the development of Big Data, the technology of Text Mining becomes mature. Thus, we analyze the price forecast model to raise the investors' rate of return. The research of Text Mining used to take the document term to analyze, but it often results in the problem with synonym and polysemy. Therefore, this research proposes a supervised learning method of sentiment pattern analysis. In addition, in order to solve the problem with training data about the supervised learning method, we mix the unsupervised learning method to carry out the subject grouping and sentimental tendency. In this study, we establish the news dataset and screen it as popular terms that are used to an unsupervised method of LDA model. The result points out that the number of news about company dropped significantly during the 2016 Taiwan president election because of the change of media sensation. Moreover, we create the sentiment dictionary that can determine the polarity of 10% neutral terms and the emotional tendency of 96% documents by mixing the NTUSD, HowNet knowledge Database and the self-expansion algorithm. Through the data visualization, the result shows that the curve of DIF-MACD is able to predict the long-term trend of 0050, while the sentiment index of the news makes a good showing in the short-term price volatility. Besides, the news sentiment index of the subjects that belong to general economy and company has about 1 to 2 day leading indicators. Eventually, we employ the Sentiment Pattern Taxonomy Model(PTM) in Chinese texts as supervised learning method and compare with VSM and SVM. The experiment result shows that PTM combined with Taiwan Weighted Stock Index is the best when its F1-Measure is up to 85%. Apart from this, we find that the sentiment index of the news in non-trading time can influence the price volatility of 0050.

Made available in DSpace on 2016-04-25T18:40:35Z (GMT). No. of bitstreams: 1 Eduardo Franco Ribeiro.pdf: 581960 bytes, checksum: 0c20561d151693cda933516c6eb0c17b (MD5) Previous issue date: 2008-11-03 / The phenomenom wich narrows and puts closer international relations around the world known as globalization has reached the development of financial markets. The Exchange Traded Funds (ETF´s) have also been involved on this process. Globalization has brought a wide exposure to international markets creating opportunities in several asset classes with lower costs and minimal investments, thus making ETF´s an extremely attractive product. This paper intends to show how ETF´s works, its creation, redemption and dealing process since such investment funds are still unknow to Brazilian investors. In addition, international portfolios will be built with ETF´s indexed cross-country equity markets, the main objective is check if a Brazilian investor may optimize his portfolio comparing international portfolios built with the brazilian asset market given by the EWZ fund / O fenômeno que estreita e aproxima as relações internacionais nos mais diversos níveis e cuja expressão globalização foi cunhada para nomeá-lo, também alcança os mercados financeiros. Os fundos de índice negociados em bolsa de valores conhecidos como ETF´s (exchange traded funds) encontram-se dentro desse processo. A possibilidade de obter exposições em diversos mercados internacionais e nas mais diversas classes de ativos a custos relativamente baixos e valores mínimos para investimento seja talvez um dos principais atrativos desse novo produto financeiro. Dessa forma, o presente trabalho pretende mostrar a criação e o funcionamento desses fundos, que ainda são pouco conhecidos no Brasil. E, a partir daí, montar carteiras internacionais, com fundos ETF indexados ao mercado de ações de diversos países, compará-las com o ativo de mercado brasileiro em dólares americanos, tentando verificar se o investidor brasileiro conseguiria otimizar a relação risco x retorno de seu portfólio por meio da diversificação com ETF´s

本研究將過去五年各ETFs基金之資料，針對Sharpe Ratio、Information Ratio、Omega Ratio、Sortino Ratio及Gain-Loss Ratio等衡量指標當作績效評比依據。 首先，分別從個別平均報酬及本利和的角度，進行ANOVA 之F檢定分析，檢視各績效指標在以一個月或三個月為衡量期間，最好(Winner)和最差(Loser)ETFs之次月(季) 平均報酬及本利和，是否存在顯著差異性。 再來，單獨利用Sharpe Ratio、Information Ratio、Omega Ratio、Sortino Ratio及Gain-Loss Ratio等指標，排序挑選每月(季)之前七檔(Winner) ETFs，除一個月為投資前間之Sharpe Ratio外，其餘各指標在次月(季)的投資績效似乎都不明顯。然而績效指標可預先發出警訊，當指標與下一個衡量期間之報酬率背離時，可當成空頭來臨前的警示燈號。 此時在研究中，思考可擬定投資策略，分別同時做多(long)其最好之七檔ETFs，及做空(short)最差之七檔ETFs，即使是經歷金融海嘯的過程，依此策略Sharpe ratio、Omega ratio、Sortino ratio或Gain-Loss ratio在絕大部分時間裡都是正報酬。 最後，迴歸分析結果顯示，要找出適合解釋ETFs報酬率能力的績效指標並不容易。並未有單一績效衡量指標具有對不同ETF皆有很好的預測能力，可見在金融市場裡，想單靠幾個績效指標來解釋ETFs基金的報酬率並不易達成。同時在研究基金績效是否具有持續性上，結論也發現以各績效衡量指標過去一個月的績效，並沒有能力去預測ETFs基金未來的價格，沒有証據可支持ETFs基金績效具有持續性。符合”所有基金績效，均為過去績效，不代表未來之績效表現”；建議機構或個別投資人買賣ETFs基金應著重於研究產經未來趨勢，而非過去績效。

L’objectif de cette thèse est d’étudier le processus social de légitimation des produits financiers en prenant comme exemple le développement des Exchange Traded Funds (ETF) en France. En dépit de la complexification de leur « dispositif de calcul » les encours des ETF ont connu une progression rapide. L'essor de ces innovations s'explique par l'effort de légitimation entrepris par les promoteurs. Contrairement à l'approche néo-institutionnaliste qui se focalise sur la dimension cognitive, nous partons des interdépendances pour comprendre le processus social de légitimation. Nous étudions ce processus en combinant une enquête ethnographique fondée sur des entretiens semi-directifs et des observations non-participantes avec une analyse du réseau social d'échange d'information entre les marketeurs, les traders, les journalistes, les investisseurs et les académiques. Nous montrons qu'il existe une « niche sociale » où des acteurs-promoteurs coopèrent entre eux. Ils relaient, auprès des diffuseurs d'information que sont les journalistes, un discours fondé sur les dimensions positives de leur innovation. Ils s'appuient, pour cela, sur le « contrôle social » qu'ils exercent sur la presse économique. Enfin, ces promoteurs s'associent à l'institut de recherche en finance Edhec-Risk afin de fonder la légitimité de leurs produits sur le statut académique de ce partenaire. / The focus of this thesis is to study the social process of legitimizing financial products, focusing primarily on Exchange Traded Funds (ETFs) in France. Despite the complex nature of their « calculative device », ETFs have grown rapidly. This can be explained through promoters’ legitimizing efforts. Unlike neo-institutional theory which focuses on the cognitive dimension, this will emphasize the role of interdependencies and relationships in studying legitimizing efforts. This study is based on ethnographic research with semi-structured interviews and non-participant observations during conferences. We combine this qualitative research method with quantitative analysis of the network of information exchanged between marketers, traders, journalists, investors, regulators and academics. We evidence the existence of a « social niche » where competitors cooperate. The promoters exert social control on financial journalists to relay the positive aspects of their financial products. They also collaborate with EDHEC-Risk Institute to benefit from its academic status and gain more legitimacy.

Submitted by Yuri Pereira (yurimedeiros_@hotmail.com) on 2016-09-01T15:24:06Z No. of bitstreams: 1 Dissertação YURI PEREIRA.pdf: 507288 bytes, checksum: b86dbb4b5f173ac7d43a83d591ab6a7b (MD5) / Approved for entry into archive by Renata de Souza Nascimento (renata.souza@fgv.br) on 2016-09-01T19:29:40Z (GMT) No. of bitstreams: 1 Dissertação YURI PEREIRA.pdf: 507288 bytes, checksum: b86dbb4b5f173ac7d43a83d591ab6a7b (MD5) / Made available in DSpace on 2016-09-01T19:33:44Z (GMT). No. of bitstreams: 1 Dissertação YURI PEREIRA.pdf: 507288 bytes, checksum: b86dbb4b5f173ac7d43a83d591ab6a7b (MD5) Previous issue date: 2016-08-05 / In Øksendal (1998), we can see the derivation of a classical stochastic optimization between an asset, or a class of assets, risky and other risk-free. But, after the decision of which portion of the resources to allocate in the risky investment class, questions arise about how would the division of the resources between the assets that comprise it. We assume that some investor choose to invest in two risky assets and, following the classic studies of portfolio stochastic optimization, mainly by Øksendal, the proposal is to introduce a new technique of trading consisting in recurrent rebalancing approach stochastic optimization investments with risk. Following the short-term concept provided by Ang, Hodrick, Xing and Zhang (2006) for the stock market, it was considered a sequence of short rebalancing time horizons and, at the beginning of each period, the parameters are recalculated and a new optimal control is established. By adopting this technique, the volatilities of the assets constituting the portfolio are recalculated and, therefore, it is a proxy to solution of the heteroscedasticity problem. Also noteworthy, being something new in literature, the fact of having been derived from an optimal control for a portfolio containing two investments with risk. The stochastic optimization procedure was similar to that adopted by Øksendal, namely, the application of the Hamilton-Jacobi-Bellman theorem to transform the problem of minimizing the cost functional a partial differential equation known as HJB equation, in reference to the authors. The steps followed by Øksenal are the same for us, from the optimization’s point of view, and are well summarized by Ross (2008). / Em Øksendal (1998), podemos ver a derivação de um modelo clássico de otimização estocástica entre um ativo, ou classe de ativos, com risco e outro sem risco. Mas, após a decisão do quanto alocar na classe de investimento com risco, ficou o questionamento sobre como ficaria a divisão dos recursos entre os ativos que a compõem. Partimos do princípio que determinado investidor optou por escolher investir em dois ativos com risco e, seguindo os estudos clássicos de otimização estocástica de portfólio, principalmente o promovido por Øksendal, a proposta é apresentar uma nova técnica de trading que consiste na abordagem de rebalanceamentos sucessivos por otimização estocástica em investimentos com risco. Seguindo a noção de curto prazo fornecida por Ang, Hodrick, Xing e Zhang (2006) para o mercado de ações, foi considerada uma sequência de horizontes curtos de rebalanceamento e, ao início de cada período, os parâmetros são recalculados e um novo controle ótimo é estabelecido. Ao adotar esta técnica, as volatilidades dos ativos que constituem o portfólio são recalculadas e, com isso, diminui-se o problema de heterocedasticidade. Também merece destaque, por ser algo novo na literatura, o fato de ter sido derivado um controle ótimo para um portfólio que contém dois investimentos com risco. O procedimento de otimização estocástica foi similar ao adotado por Øksendal, qual seja, a aplicação do teorema de Hamilton-Jacobi-Bellman para transformar o problema de minimização da funcional custo numa equação diferencial parcial conhecida como equação HJB, em referência aos autores. Os passos seguidos por Øksenal e por nós serão os mesmos, do ponto de vista de otimização, e estão bem resumidos por Ross (2008).

The subject of the diploma thesis is to evaluate the development in post-crisis years and to determine the impacts that affect the current financial investment environment in the USA but also its future and create the prerequisites for other risks, which the market can affect in the next perspective and influence the global development. On this basis, a portfolio of the mutual fund will be drawn up, according to the assignment of its manager with an alternative investment component, which can adequately complement it in the current period of low interest rates.

This study investigates how term structure dynamics of VIX futures can be exploited forabnormal returns. To be able to access volatility as a tradeable asset, the trading strategiesonly trades ETFs which are designed to replicate the movements of VIX futures index. Itis established that such ETFs are unsuitable for buy-and-hold investments because of thenegative roll yield it usually suffers, caused by the slope of the VIX term structure.Consequently, these conditions create opportunities for strategies that use direct andinverse VIX ETFs to be profitable. The study is a quantitative study that uses historicalprice data to back test three different trading strategies. The strategies are tested over theperiod 11-oct-2011 to 31-mar-2020. The authors have deliberately chosen to delimit thestudy by not testing the performance of the ETFs, not statistically test the risk-adjustedreturns and not perform a regression to calculate optimal hedge ratios for the strategies.The results from this study shows that its possible for strategies that exploit the termstructure dynamics of VIX futures to generate abnormal returns.

本文以臺灣證交所上市的前九檔槓桿及反向指數股票型基金(LETFs)作為 樣本,根據 Charupat and Miu(2014)研究方法衡量其績效。傳統衡量指數股票 型基金(ETFs)績效的方式,為單純將基金淨值報酬對指數累積報酬做簡單迴 歸,若將此方法應用在衡量 LETFs 之績效上,由於許多影響基金淨值報酬的因素 沒被分離出來,常造成迴歸結果存在嚴重偏誤,或是難以解釋。本文是第一篇研 究國內 LETFs 績效的著作,透過在迴歸式中納入複利效果、融資效果,以更精確 的方式比較分析影響 LETFs 基金淨值報酬的因素,及各 LETFs 之管理績效。本 文實證結果除了證實融資效果確實存在,也證實大部分複利效果及融資效果的理 論性質,最重要的是,顯示出追蹤上証 180 指數的三檔 LETFs 在準確複製報酬槓 桿倍數上比較傑出,而追蹤台灣 50 指數的三檔 LETFs 在基金管理效率方面有比 較優秀的表現。 / Using Leveraged and Inverse Exchange-Traded Funds (LETF) listed in the Taiwan Stock Exchange, this thesis evaluates the performance of these LETFs based on the methodology proposed by Charupat and Miu (2014). The traditional approach of performance evaluation of ETFs is to regress the fund’s net asset value (NAV) returns on the underlying index’s returns. However, such an approach fails to account for important factors, such as compounding and financing effects, that affect the NAV of the LETFs, and unavoidably leads to serious estimation biases. This is the first thesis which evaluates the performance of LETFs listed in the Taiwan Stock Exchange. By considering compounded effect and financing effect in the regression model, the proposed method is more precise and appropriate in disentangling factors that affect the performance of the LETFs. Our empirical evidence shows how compounding effect, financing costs, and management factors influence LETFs’ tracking errors. Most of all, the three LETFs tracking the SSE180 index have the best tracking ability of the underlying asset return, while the LETFs tracking the FTSE TWSE Taiwan 50 index have the best management performance among all LETFs examined in this these.

從投資組合理論與效率市場假說來看，影響投資績效之主要因素為資產配置而非選股或擇時能力，相關之實証亦發現主動式投資策略並無法獲得持續優於大盤的報酬，因此被動式投資策略將是一較佳的選擇。自先鋒集團在1976年率先推出指數基金之後，此類指數化投資商品即呈現指數化成長。而我國亦在2003年6月30日正式推出第一檔指數股票型基金「臺灣50指數股票型基金，簡稱TTT」。 本論文以投資人的角度來分析TTT之報酬率及成交量，從追蹤誤差及溢折價來觀察基金之報酬率，研究造成追蹤誤差之主要影響因子為何與實物創造或贖回機制是否能有效發揮功用，使TTT之市價可貼進基金淨值，以免如同封閉型基金大都折價交易，而影響投資人之獲利率。此外，更進一步比較TTT與其他指數化商品之差異性以供投資人選擇投資標的時參考。 本研究發現影響追蹤誤差之因素主要為指數成份股之增刪、公眾流通係數之調整與現金股利的發放。雖然在短期下上述因素會影響基金報酬，但長期而言基金報酬與指數相當，甚或更為略高，故在衡量長期績效時臺灣50指數可做為一良好的報酬指標。而從溢折價來分析時，由於市場交易機制，使得TTT自上市以來之流動性不足，市場效率性不高，溢折價幅度較大。此點與成交量之分析結果一致，研究發現TTT之成交量並未因出現套利機會而顯著增加，投資人買賣TTT主要在於避險或投機需求。 雖然現階段我國尚未出現指數基金，但若未來有出現類似之商品時，在目前我國停徵證券交易所得稅的環境下，投資金額的大小將非決定選擇指數基金或TTT之關鍵因素，投資期間與「質」的因素才是主要關鍵。當投資期間愈長，TTT之高交易成本的影響程度將降低，其可在盤中隨時買賣及低追蹤誤差等「質」方面的優勢將提高，TTT將相對較具吸引力。而臺灣50指數期貨由於交易成本及流動性風險較高，因而投資人較不偏好操作臺灣50指數期貨。

The goal of this diploma thesis is to analyze the gold market and provide readers with the necessary information and context having an impact on the price of gold. The thesis consists of three chapters. First one deals in general with the commodity market and introduces the readers to commodity exchange issues such as trading commodities in commodity exchanges, motives of commodity trading as well as the specific characteristics of commodities. Second one concerns the detailed analysis of commodity investment tools that investors might use when they feel like getting an exposure to price movements of commodities. The last chapter gears towards an analysis of the gold market in today's super globalized world and depicts what fundamental factors have an impact on the price of gold. At last, I shall summarize existing pieces of knowledge and cast light on further gold price movements.