[[alternative]]權益型不動產投資信託波動性預測與特性

博士[[abstract]]本論文著重於權益型不動產投資信託波動性預測與特性，共包含三個部份。第一部份為「美國權益型不動產投資信託之波動性預測」、第二部份為「權益型不動產投資信託與股市、利率之敏感性」與第三部份為「權益型不動產投資信託風險溢酬與不對稱波動性」。將此三部份的內容簡述如下。 第一部份針對美國權益型不動產投資信託比較WCARRX模型與ECARRX模型進行波動性預測。實證結果簡述如下：1.檢驗變幅衝擊長期的穩定條件知悉美國權益型不動產投資信託資料較符合WCARRX模型。2.本文檢驗長期利率與股價指數將正向衝擊美國權益型不動產投資信託。此外，西德州原油指數對美國權益型不動產投資信託是不顯著，其隱含美國權益型不動產投資信託無法呈現有效的通貨膨脹避險。3.樣本外之預測支持WCARRX模型優於ECARRX模型。由上述之預測評估可以觀查到WCARRX模型較優於ECARRX模型波動性預測。 第二部份本文主要研究使用ARJI模型去檢驗美國與日本權益型不動產投資信託市場，對股市與長短期利率敏感性之動態分析。實證結果簡述如下：1.權益型不動產投資信託報酬與兩國之股價指數報酬皆呈現顯著之正相關，其隱含權益型不動產投資信託行為比較像普通股(小型股)標的而非標的不動產或債券。因此，權益型不動產投資信託給投資人好的分散風險收益。2.本文發現權益型不動產投資信託報酬與10年期政府公債報酬正相關，其隱含利率上升可能反應好的經濟成長與預期通貨膨脹，進而推升不動產價格。3.權益型不動產投資信託報酬與3個月期政府公債報酬皆無顯著性相關。最後證實兩國權益型不動產投資信託報酬皆存在高度之波動性叢聚現象。 第三部份本文使用GJR-ARJI模型去檢驗美國、澳洲與日本權益型不動產投資信託之超額報酬與市場投資組合之超額報酬。實證結果簡述如下: 1.權益型不動產投資信託之超額報酬與市場風險溢酬呈現相同方向關係。其隱含基於權益型不動產投資信託多樣化之特性權益型不動產投資信託指數與股票指數呈正相關。我們建議一個理性投資人在股市過熱時選擇權益型不動產投資信託指數諸如基金或債券避險。2.本文發現基於過去好消息和壞消息衝擊造成不對稱效果。實證結果證明權益型不動產投資信託超額報酬與市場風險溢酬皆存在高度之波動性叢聚現象。因此，本文建議使用GJR-ARJI模型去捕捉與理解美國、澳洲與日本權益型不動產投資信託之超額報酬特性與避免不正確的財務與經濟之決定。[[abstract]]The purpose of this dissertation is to contribute to the literature on volatility forecasting and characteristics of Equity REITs which comprises three parts. The first part is entitled “Forecasting Volatilities for U.S. Equity REITs”, the second part is named “Stock and Interest Rate Sensitivity of Equity REITs”, and the last one is “Risk Premium and Asymmetric Volatility of Equity REITs”. A brief introduction of these three parts can be summarized as follows: The first part compares the WCARRX model with the ECARRX model in forecasting financial volatilities for U.S. Equity REITs. The empirical results are summarized as follows. First, we indicate the persistence of range shocks that U.S. Equity REITs data seem to support a Weibull alternative over the null of an exponential distribution. Secondly, this dissertation investigates that the long-run interest rate and stock market have positive shock with the U.S. Equity REITs. Furthermore, this dissertation finds that there is not statistically significant in the case of the West Texas crude oil Index, which implies that the U.S. EREITs do not represent effective inflation hedges. Third, out-of-sample volatility forecasts give rise to almost unanimous support for the WCARRX model over the ECARRX model. As a result of the above forecast evaluations, it is obvious that the WCARRX model does provide sharper volatility forecasts than the ECARRX model. In the second, the main study of this dissertation uses the ARJI model to examine the U.S. and Japan EREITs markets, which with the dynamic analysis of stock and long-term and short-term interest rate sensitivity. The empirical results have summarized as follows. First, our results show that there exists a positive correlation between the target EREITs returns and the market index. This implies that EREITs behave more like common stocks (small stocks) than the underlying real estate or bond. Furthermore, the results show that REITs, in particular EREITs, offer investors good diversification benefits. Secondly, this dissertation finds a positive correlation between the target EREITs returns and the yield of 10-year Treasury notes in recent years. This implies that an increase in interest rates may reflect stronger economics growth, higher inflationary expectations, and upward pressure on real estate prices. Third, this dissertation finds that there is not statistically significant between the target EREITs returns and the yield of 3-month T-bills. Finally, empirical results demonstrate that the return of U.S. EREITs and Japan EREITs indices have highly volatility clustering phenomenon. The last part utilizes the GJR-ARJI model to examine the daily excess returns of the EREITs index and the daily excess returns of market portfolios in the U.S., Australia and Japan EREITs indices. The empirical results have summarized as follows. First, the excess EREITs returns and the market risk premium are related in that they move in the same direction. This implies that the international EREIT indices and stock indices are positive correlated based on the EREITs’ diversified characteristics. We suggest that a relational investor can choose a REIT index to replace an overreacting stock index such as a hedge fund or a bond. Secondly, this dissertation finds evidence of a strong asymmetric effect with respect to the impact of past good and bad news. The empirical results demonstrate that the excess EREIT returns and market risk premium exhibit jump phenomena. Hence, this study suggests using the GJR-ARJI model investigate the excess return concept to capture and comprehend the true features of the EREITs for the U.S., Australia and Japan, and thus avoids incorrect financial and economic decisions.[[tableofcontents]]TABLE OF CONTENTS Page ACKNOWLEDGEMENT i ABSTRACT IN CHINESE ii ABSTRACT IN ENGLISH iv LIST OF TABLES ix LIST OF FIGURES x PART I 1 Forecasting Volatilities for U.S. Equity REITs ABSTRACT 2 CHAPTER 1. Introduction 3 1.1 Motivations and Objectives 3 1.2 Flow Chart 5 2. Literature Review 6 3. Econometric Methodology 11 3.1 The Weibull Conditional Autoregressive Range (WCARR) model 11 3.2 Out-of-Sample Volatility Forecasting Comparison 15 4. Data Description and Empirical Results 16 4.1 Data Description 16 4.2 Estimation Results 17 4.3 Out-of-Sample Volatility Forecast Comparison 22 5. Concluding Remarks 25 BIBLIOGRAPHY 27 PART II 34 Stock and Interest Rate Sensitivity of Equity REITs ABSTRACT 35 CHAPTER 1. Introduction 36 1.1 Motivations and Objectives 36 1.2 Flow Chart 39 2. Literature Review 40 3. Econometric Methodology 45 3.1 ARJI Model 45 4. Data Description and Empirical Results 49 4.1 Data Description 49 4.2 Empirical Results 52 5. Conclusions 55 BIBLIOGRAPHY 57 PART III 62 Risk Premium and Asymmetric Volatility of Equity REITs ABSTRACT 63 CHAPTER 1. Introduction 64 1.1 Motivations and Objectives 64 1.2 Flow Chart 68 2. Literature Review 69 3. Econometric Methodology 73 3.1 GJR-ARJI Model 73 4. Data and Empirical Results 77 4.1 Data Description 77 4.2 Empirical Results 78 5. Conclusions 87 BIBLIOGRAPHY 89 LIST OF TABLES Page PART I Table I.1. Summary Statistics for the Returns and Ranges of Daily U.S. EREITs Index 16 Table I.2. Estimation of the ECARR Models Using Daily U.S. EREITs Index With Exponential Distribution-The ECARR(1,1) Model 20 Table I.3. Estimation of the WCARR Models Using Daily U.S. EREITs Index With Weibull Distribution-The WCARR(1,1) Model 21 Table I.4. Out-of-Sample Forecasting of ECARR AND WCARR Forecast from EREITs 23 PART II Table II.1. Descriptive Statistics of Daily Returns 51 Table II.2. Estimation of the ARJI Model with two target U.S. and Japan EREITs Indices 54 PART III Table III.1. Descriptive Statistics of daily excess equity REIT Index returns 79 Table III.2. Estimation of the ARJI Model with the Excess Return Model 82 Table III.3. Estimation of the GJR-ARJI Model with the Asymmetric Excess Return Model 86 LIST OF FIGURES Page PART I Figure I.1 The U.S. EREITs index in level and daily returns 16 PART II Figure II.1 The two target EREITs indices in levels, daily returns and daily returns density 50 PART III Figure III.1. The three target EREITs indices & stock indices in levels……………… 78[[note]]學號: 894490118, 學年度: 9