金融市場中個別資產的風險感染效果越趨嚴重，使得傳統資產配置理論下的投資組合面臨極大的虧損。有鑑於此，若能在投資組合模型中納入考量此種擴散效果，將可更加分散風險以增進投資組合的效率性，並進一步降低投資組合面臨極端虧損的可能性。因此，要如何納入此一風險擴散效果，以在良好的風險控管下進行資產配置，將可能遭受的損失降至最低，是本論文主要探討的問題。 本研究延伸Adrian, Brunnermeierz (2009) CoVaR的概念，納入考量系統性風險因素，透過CoVaR模型衡量系統性風險擴散時，造成個別標的資產報酬率變動的程度，並將Markowitz的效率前緣加以改良，建構更具效率性的Mean-CoVaR資產配置模型，以計算新的最適配置權重與最適投資組合。此外，本研究也就Mean-CoVaR資產配置模型與傳統Markowitz(1952)所提出的Mean-Variance模型進行探討與比較。 綜合本研究之實證結果，Mean-Variance模型雖然能使投資組合報酬率的波動度最小，但在面臨極端系統性風險下，其績效表現卻不如Mena-CoVaR模型所建構出的投資組合；因此，在傳統的Mean-Variance模型下，若能以CoVaR取代Variance所建構出新的Mean-CoVaR投資組合模型，納入大盤風險可能的擴散效果下，將可有效降低投資組合在大盤崩跌時的虧損程度，以維持較佳的投資績效。

Um dos objetivos deste estudo é testar modelos de precificação de ativos financeiros, especialmente o de risco downside de Ang et al. (2006), em todas as ações da Bovespa, para o período que se estende de janeiro de 1999 a julho de 2012. Para atingi-lo, aplica-se o método de regressões Fama e MacBeth (1973) com retornos um período à frente. A quase totalidade dos modelos testados é rejeitada, inclusive a existência de um eventual prêmio para o risco downside. A exceção é o modelo que inclui com o beta tradicional e o seu quadrado, o que permite rejeitar o CAPM devido a não linearidade no risco de mercado. A relação existente entre o beta e o retorno das ações seria positiva até beta igual a 0,642 e depois negativa. Outra meta desta dissertação é comparar as estimações condicionais às não condicionais do modelo CoVaR de Adrian e Brunnermeier (2011) para as 16 ações da Bovespa utilizadas por Almeida et al. (2012), que obtiveram apenas estimações não condicionais para o Brasil em um período semelhante. Os resultados daqui mostram uma baixa e não estatisticamente significante correlação com os de Almeida et al. (2012). Para este estudo, tem-se que as duas formas de calcular o CoVaR são similares para o teste de estresse, mas não para o risco sistêmico. / This research pursues as an objective to test cross-sectional returns of some asset pricing models, specially the downside risk suggested by Ang et al. (2006). To accomplish this goal, all the Brazilian Bovespa\'s stocks are used, from January 1999 to July 2012, in one month forward returns Fama-MacBeth regressions. Not only the downside risk model is rejected: almost all models, including the traditional CAPM and versions of the 3 factors Fama-French. A nonlinear CAPM (beta and beta squared) is the exception in the universe of tested models, which produces the best predictions and a positive relationship between betas and forward returns until beta equals 0,642, after this value, the relationship becomes negative. Another issue followed by this study is to compare conditional estimates of the CoVaR model of Adrian and Brunnermeier (2011) with the unconditional ones for the sixteen stock used by Almeida et al. (2012) unconditionally estimates. The results show low and not statistically significant correlation with Almeida\'s estimates. For the sample used here, comparing the conditional and the unconditional methodologies suggests a great similarity for the stress test, but not so close results for the systemic risk.

In this paper we use the conditional Value at Risk (CoVaR) and CoVaR variation (ΔCoVaR) proposed by Adrian and Brunnermeier (2008, 2011, 2016) to estimate the Peruvian stock market risk (through the IGBVL) conditioned on the international financial market (given that the S&P500) and conditioned on three of the main commodities exported by Peru: copper, silver and gold. Moreover, the CoVaR measures are compared with the VaR of the IGBVL to understand the differences using conditional and unconditional risk measure estimators. The results show that both CoVaR and ΔCoVaR are useful indicators to measure the Peruvian stock market risk. / En este trabajo utilizamos el Valor en Riesgo condicional (CoVaR) y la variación CoVaR (ΔCoVaR) propuestos por Adrian and Brunnermeier (2008, 2011, 2016) para estimar el riesgo bursátil peruano (a través del IGBVL) condicionado en el mercado internacional (dado por el índice S&P500) y condicionado en tres de los principales comodities exportados por el Perú: cobre, plata y oro. Además, las medidas CoVaR son comparadas con el VaR del IGBVL para entender las diferencias al utilizar medidas de riesgo condicionales e incondicionales. Los resultados muestran que ambas medidas CoVaR and ΔCoVaR constituyen indicadores útiles para estimar el riesgo bursátil peruano.

Um dos objetivos deste estudo é testar modelos de precificação de ativos financeiros, especialmente o de risco downside de Ang et al. (2006), em todas as ações da Bovespa, para o período que se estende de janeiro de 1999 a julho de 2012. Para atingi-lo, aplica-se o método de regressões Fama e MacBeth (1973) com retornos um período à frente. A quase totalidade dos modelos testados é rejeitada, inclusive a existência de um eventual prêmio para o risco downside. A exceção é o modelo que inclui com o beta tradicional e o seu quadrado, o que permite rejeitar o CAPM devido a não linearidade no risco de mercado. A relação existente entre o beta e o retorno das ações seria positiva até beta igual a 0,642 e depois negativa. Outra meta desta dissertação é comparar as estimações condicionais às não condicionais do modelo CoVaR de Adrian e Brunnermeier (2011) para as 16 ações da Bovespa utilizadas por Almeida et al. (2012), que obtiveram apenas estimações não condicionais para o Brasil em um período semelhante. Os resultados daqui mostram uma baixa e não estatisticamente significante correlação com os de Almeida et al. (2012). Para este estudo, tem-se que as duas formas de calcular o CoVaR são similares para o teste de estresse, mas não para o risco sistêmico. / This research pursues as an objective to test cross-sectional returns of some asset pricing models, specially the downside risk suggested by Ang et al. (2006). To accomplish this goal, all the Brazilian Bovespa\'s stocks are used, from January 1999 to July 2012, in one month forward returns Fama-MacBeth regressions. Not only the downside risk model is rejected: almost all models, including the traditional CAPM and versions of the 3 factors Fama-French. A nonlinear CAPM (beta and beta squared) is the exception in the universe of tested models, which produces the best predictions and a positive relationship between betas and forward returns until beta equals 0,642, after this value, the relationship becomes negative. Another issue followed by this study is to compare conditional estimates of the CoVaR model of Adrian and Brunnermeier (2011) with the unconditional ones for the sixteen stock used by Almeida et al. (2012) unconditionally estimates. The results show low and not statistically significant correlation with Almeida\'s estimates. For the sample used here, comparing the conditional and the unconditional methodologies suggests a great similarity for the stress test, but not so close results for the systemic risk.

本研究欲以分量迴歸的方法估計出台灣上市櫃金融控股公司的VaR、CoVaR及其對台灣金融市場的風險溢出，做為總體審慎監理原則下具有抗景氣特色之風險衡量參考指標。我們亦透過金控公司間之CoVaR，觀察金控公司間風險交互影響程度，盼可提供各金控公司做為個體審慎監理原則下風險管理之參考指標。 本研究包含四大特色：一、運用前期市場資料可估計下期含有條件、共變、傳染、貢獻等特性之風險值，也就是CoVaR；二、透過各家金控對市場之∆CoVaR可觀察各金控公司系統風險貢獻程度差異；三、可觀察金控公司間相互交叉影響程度；四、運用金融機構特性預測未來系統風險。 本研究以信用利差、長短期利差、流動性利差、匯率變動、加權指數報酬、隱含波動度變動、金控股價報酬等市場資料，透過分量迴歸估計損失機率為1%及5%之台灣金融控股公司VaR及CoVaR，並計算市場風險溢出─∆CoVaR研究各金融機構對系統風險之邊際貢獻。且以槓桿比率、市值帳面比、相對規模及資產負債不對稱比例等金融機構特性相關變數預測未來∆CoVaR，做為總體審慎監理原則下之風險管理參考指標。 本研究結果發現對台灣金融市場系統風險溢出貢獻較大的為玉山金、中信金、台新金及國泰金；國票金、永豐金、第一金及元大金則為系統風險溢出貢獻較低者。預測結果部分發現損失機率為1%時，以預測未來兩季之∆CoVaR效果較佳，預測損失機率為5%時則以預測未來三季之∆CoVaR效果較佳，顯示資料對不同的尾端損失機率分配影響顯現時間也不相同。 / In this thesis, we intend to estimate Taiwanese financial holding company’s VaR, CoVaR and risk spillover to Taiwan financial market, and apply these to macroprudential risk management. In addition, we intend to develop crossover CoVaR between financial holding companies, offering risk management referral benchmark under microprudential principle to those companies. There are four features in this thesis. First, we use previous market data to estimate the conditional, comovement, contagion, and contributing VaR - CoVaR. Second, by ∆CoVaR of the institutions to the market, we can observe the holding companies’ systematic risk contribution. Third, we can observe the crossover effect of the holding companies. Last, we could use the characteristics of the institutions to predict future systematic risk. We particularly use credit spread, slope of yield curve, liquidity spread, change of exchange rate, return of market stock index, change of implied volatility and holding company’s stock price, by quantile regression, to predict the VaR and CoVaR of Taiwan’s holding companies when the probability to loss is 1% and 5%. Then we calculate market systematic risk spillover, ∆CoVaR, to observe the marginal systematic risk contribution of the institutions. Moreover, we use leverage, market-to-book ratio, relative size and maturity mismatch to predict forward ∆CoVaR, offering a reference to macroprudential risk management. Our empirical results show that in Taiwan financial market, the top four systematic risk contributors of holding companies are Esun Financial Holding, Chinatrust Financial Holding, Taishin Financial Holding and Cathay Financial Holding; the smallest ones are Waterland Financial Holding, Sino Financial Holding, First Financial Holding and Yuanta Financial Holding. We also find out that when loss probability is 1%, predicting ∆CoVaR after two seasons is better; when loss probability is 5%, predicting ∆CoVaR after three seasons is more significant. It shows that when the tail is different, the effect time is also different.

With a background in EU regulations and an increased interest in Environmental, Social, and Governence (ESG) policies in companies when investing, this thesis considers the individual contributions to market risk in portfolios by different ESG parameters. It explores two different methods to examine if there are effects consistent across the whole Nordic markets, and the possibility to express any effects within portfolios in a clear way. It uses the OMXNORDIC index as the market index and two different fund portfolios as example portfolios, one of which is an article 9 fund. The quantile regression approach does not show any consistent effects across the whole Nordic market from any ESG parameter explored. It does however make for a clear way to present the effects on the portfolio level for each ESG parameter. The employed Copula approach does show some consistent difference between the ESG parameters for the market and in portfolios, as well as differences between the portfolios. Both of the explored methods should allow for comparisons between, and reports on, fund portfolios which would improve the ESG analyses of funds. / Mot bakgrund av EU-lagstiftning och ett ökat intresse i företags förhållning till Environmental, Social, och Governence (ESG) frågor, så utforskar den här uppsatsen ESG-faktorers bidrag till marknadsrisk i fondportföljer och på den nordiska marknaden. Uppsatsen använder två olika metoder för att undersöka om det finns potentiella konsekventa effekter på den Nordiska aktiemarknaden, och möjligheten att presentera resultat på portföljnivå på ett tydligt sätt. OMXNORDIC används som marknadsindex, och två olika fondportföljer används som exempelportföljer, varav en är en artikel 9 fondportfölj. Quantile regression-metoden visar inte på några konsekventa effekter över hela den nordiska marknaden, för någon av ESG-parametrarna. Däremot så resulterar metoden i ett tydligt sätt att presentera påverkan av ESG-parametrarna på portföljnivå. Copula-metoden som används visar på några konsekventa skillnader mellan ESG-parametrar, både för marknaden och i fondportföljerna, samt skillnader mellan portföljerna i sig. Båda metoderna lämpar sig till att jämföra och bygga rapporter på fondportföljer, vilket borde leda till bättre ESG-analyser av fonder.

由於 2007-2009 年金融風暴的發生 , 使得系統風險的研究受到相當大的關 注 , 而此論文也將探討台灣金融業的現況 。 我們根據Adrian et al.(2016) 、 Acharya et al.(2012) 以及 Brownlees et al.(2012)所提出的Delta CoVaR 、 MES 以及 SRISK 等系統風險衡量指標 , 估算台灣金控系統風險的大小 , 以及評 斷台灣系統風險重要金控的排序 。 透過時間序列及橫斷面的分析 , 我們更 將風險趨勢分群 , 或是從相關風險指標來作為監督機構或投資大眾參考的 早期警訊 。 最後 , 我們亦透過追蹤資料模型 , 找出系統風險重要的解釋變數 , 並分析變數的可能影響效果 。 / After the Financail Crisis of 2007-2009, there have been rich research about systemic risk analysis, and this work focus on financial industry in Taiwan. According to Adrian et al.(2016)、 Acharya et al.(2012)and Brownlees et al.(2012), we consider four measures for systemic risk,they are MES、SRISK、Delta CoVaR-DCC and CoVaR-Quantile. We demonstrate how to compare four different measures , and display the ranking of the Systemically Im- protant Financial Institutions (SIFs) based on the resulting SRISK, for Taiwanese holding companies. Finally , we also dicuss the individual and macroeconomic effects on systemic risk by using panel data regression .

近年來金融性風暴頻傳，導致全球資金快速移動，許多國家的股匯市因此產生劇烈波動，台灣即是其中之一。有鑑於台灣股匯市的波動，部分投資者開始採用股價與匯率的相關性進行未來走勢預測，並建構策略進行交易，但中央銀行一再宣稱台灣股匯市間不存在實質的相關性，並提醒投資大眾不要因錯誤認真而遭逢重大損失，因此本研究的主要目的在於分析極端事件下台灣股匯市的關聯性。 本研究採用新的風險評估方法「CoVaR」進行分析，其定義為在其他市場發生特殊事件下目標市場的最大可能損失，而CoVaR與VaR的主要差別在於其考慮了其他市場的外溢效果，因此更能充分反映極端事件下的真實風險值。本文採用1993年至2011年的台灣加權股價指數和美元兌台幣匯率日資料，經由實證分析後主要有三大發現:一、美元匯率報酬臨界值與股價指數報酬率呈現負相關，股價指數報酬臨界值也與美元匯率報酬率呈現負相關；二、整體而言，股市多方比空方承受更多的風險，新台幣持有者比美元持有者承受更多的風險；三、股市對匯市的外溢效果較匯市對股市來的強烈。此外，台灣股匯市若採用新風險評估方法CoVaR進行風險值估算，將可以發現其較傳統VaR高出兩成至七成，由此可知台灣股匯市若處於極端事件下，將產生嚴重的風險外溢現象。

In the new context of major public emergencies, this paper will mainly study the measurement and influencing factors of systemic risks in Chinese financial institutions based on three dimensions: overall situation, industries, and institutions. First, it uses the DTW-MST network model to describe the dependence structure between financial institutions and between industries. It explores important institutional nodes of risk dependence from a network perspective. Then, it uses the time-varying Copula-CoVaR model to measure financial institutions' and industries' risk spillover effect on the whole financial system and analyze the characteristics and differences of risk spillover. Last, it uses the panel regression model to study the influencing factors of the risk spillover effect of financial institutions and explore the sources of systemic risks. The results show that: (1) Industrial Bank (CIB), Changjiang Securities (CJSC), and China Pacific Insurance (CPIC) are the central nodes of the banking, securities, and insurance industries, respectively. (2) The risk spillover effect is characterized by a significant asymmetry and thick tail, and negative news has a greater impact on the risk spillover effect. (3) The value at risk (VaR) and volatility of financial institutions have a significant positive correlation with the risk spillover effect, while the size of financial institutions has a significant negative correlation with the risk spillover effect. / Business Administration/Finance

Le risque systémique a joué un rôle clé dans la propagation de la dernière crise financière mondiale.Un grand nombre de mesures de ce risque ont été développées pour évaluer la contribution d’une institutionfinancière au risque de l’ensemble du système. Toutefois, de nombreuses questions concernantles capacités de ces mesures à identifier les institutions financières d’importance systémique (SIFIs) ontété soulevées puisque le risque systémique possède de multiples facettes et certaines d’entre elles sontdifficiles identifier, telles que les similitudes entre institutions financières.L’objectif général de cette thèse en finance est donc (i) de proposer une solution empirique pour identifierles SIFIs au niveau nationale, (ii) de comparer théoriquement et empiriquement différentes mesures durisque systémique et (iii) de mesurer les changements d’expositions au risque des banques.Tout d’abord, le chapitre 1 propose un ajustement de trois mesures de risque systémique basées sur desdonnées de marchés et conçues dans un cadre international, afin d’identifier les SIFIs au niveau national.Ensuite, le chapitre 2 introduit un modèle commun dans lequel plusieurs mesures du risque systémiquesont exprimées et comparées. Il y est théoriquement établi que ces mesures de risque systémique peuventêtre exprimées en fonction de mesures traditionnelles de risque. L’application empirique confirme cesrésultats et montre que ces mesures ne sont pas capables de saisir la nature multidimensionnelle durisque systémique. Enfin, le chapitre 3 présente la méthodologie appelée Factor Implied Risk Exposures(FIRE) permettant de décomposer une variation de la mesure de risque d’une banque en deux éléments,le premier représentant la volatilité de marché et le second correspondant à l’exposition au risque de labanque. Ce chapitre illustre empiriquement que les changements d’expositions au risque sont corréléspositivement entre les banques, ce qui est cohérent avec le fait que les banques présentent des similitudesdans leurs prises de positions sur le marché. / Systemic risk has played a key role in the propagation of the last global financial crisis. A large number ofsystemic risk measures have been developed to quantify the contribution of a financial institution to thesystem-wide risk. However, numerous questions about their abilities to identify Systemically ImportantFinancial Institutions (SIFIs) have been raised since systemic risk has multiple facets, and some of themare difficult to gauge, such as the commonalities across financial institutions.The main goal of this dissertation in finance is thus (i) to propose an empirical solution to identifydomestic SIFIs, (ii) to compare theoretically and empirically different systemic risk measures, and (iii)to measure changes in banks’ risk exposures.First, chapter 1 offers an adjustment of three market-based systemic risk measures, designed in a globalframework, to identify domestic SIFIs. Second, chapter 2 introduces a common framework in whichseveral systemic risk measures are expressed and compared. It is theoretically shown that those systemicrisk measures can be expressed as function of traditional risk measures. The empirical application confirmsthese findings and shows that these measures fall short in capturing the multifaceted nature of systemicrisk. Third, chapter 3 proposes the Factor Implied Risk Exposures (FIRE) methodology which breaksdown a change in risk disclosure into a market volatility component and a bank-specific risk exposurecomponent. This chapter empirically illustrates that changes in risk exposures are positively correlatedacross banks, which is consistent with banks exhibiting commonality in trading.