臺灣大學: 土木工程學研究所荷世平鄭涵因Cheng, Han-YinHan-YinCheng2013-04-012018-07-092013-04-012018-07-092012http://ntur.lib.ntu.edu.tw//handle/246246/255519過去文獻有關市場發展過程討論最多的莫過於形成市場泡沫、崩跌時期的原因，對於市場循環發展過程缺乏有系統的一致架構，導致其他市場發展時期含有的重要資訊可能被忽略。資產市場報酬波動不對稱性描述不同方向的價格趨勢會造成不同程度的波動，進一步強調價格動態對形成市場波動特徵的重要性。本研究假設不同型態的短期價格趨勢型態，有著不同的市場波動特徵，且其中隱含著牛、熊市下異質信念交易者的投資策略。為了討論市場循環中各階段的市場特徵，我們使用數據分析法(Dating Algorithm)將美國資產市場資料以價格趨勢分段，並以GJR-GARCH波動不對稱報酬時間序列模型估計各資料段的模型參數。研究最後提出一個系統化的市場機制完整描述市場循環中各階段異質信念投資者投資策略互動情形以及市場特徵的成因。There are some special market stages during market cycle, while research in the past focus on some extreme situation in the market without a systematic description for the development of the market which may result in omitting important information contained in the stages they have ignored. The well-known characteristic of equity markets, asymmetric volatility, which means opposite directions of price movement cause distinct level of volatility further emphasizes the vital role price dynamics plays in forming characteristics of volatility. In our research, we assume that different kinds of price dynamics, market trend, in the short run under bull and bear market phases have distinct characteristics of volatility implying trading strategies of heterogeneous beliefs. In order to explore the features of each market stage, we use dating algorithm to divide the U.S. market data series into sub-periods according to price trend and analyze the volatility characteristics of each sub-period data series by GJR-GARCH model piecewisely. We come up with a novel systematic way to describe the market mechanism which contains each kind of investor’s trading strategies and their influences on price dynamics to form market stages that with different volatility characteristics under market cycle.2232502 bytesapplication/pdfen-US市場循環波動異質信念GJR-GARCH不對稱Market cyclevolatilityheterogeneous beliefasymmetrythesishttp://ntur.lib.ntu.edu.tw/bitstream/246246/255519/1/ntu-101-R98521730-1.pdf