本文建構全球、亞洲地區及亞洲四小龍投資人的情緒指標，驗證其對於當地市場價格及波動的預測能力、檢視跨國投資人情緒之外溢傳導效果，並比較亞洲四小龍市場之持續買進(hold-more)、價格壓力(price-pressure)、Friedman 或空間創造(create-space)效果。本文相對既有文獻的創新之一，在於同時考量投資人的高、低情緒狀態與市場多、空結構，探討情緒對於市場報酬及波動造成的影響。再者，本文擬透過馬可夫模型不同狀態的機率計算，描述投資人情緒變化過程，以評估投資人情緒改變對於市場價格之影響。本計劃另一研究重點，擬在DCC-BC-MVHRs (Dynamic Correlation Coefficient Basis-Convergence MVHRs)模型架構下，檢視投資人情緒對於現貨報酬與基差行為之影響，觀察樣本內和樣本外之短期、中期和長期避險績效的提升。此外，並進一步區分投資人情緒為高情緒、低情緒、避險情緒和投機情緒等，據以比較不同型態投資人情緒在股票、外匯、商品、及能源市場等不同結構市場避險績效的差異。其次，並應用到風險值估算，評估情緒對於風險值估計效能是否有效提升。最後，針對投資人情緒變動與市場報酬間可能存在之尾部極值相關特性，本計畫擬針對兩者間的變動行為，透過copula 函數作探討，補充既有文獻的不足。 The paper examines the sentiment predictability of Four Asian Tigers on market price and volatility by constructing global, regional and local sentiment indicators. We further investigate the spillover effect of sentiment across countries and also verify the effects of hold-more, price-pressure, Friedman and create-space within Four Asian Tigers. Comparing with prior literature, we simultaneously take the regime of investor sentiment and market structure into account when analyzing the effect of sentiment on stock return and volatility. Therefore, one of our main contributions is to characterize the market dynamics by adopting the Markov-Switching model to distinguish the effect of time varying sentiments on market price. In addition, with DCC-BC-MVHRs (Basis-Convergence MVHRs) model, we examine the effect of investor sentiment on stock return and basis, and observe the short-term, mid-term and long-term hedge performance of in- and out-of sample. We then compare the hedging performances for stocks, exchange rates, energy and commodities markets under high, low, hedge and speculative sentiments and evaluate the effectiveness of sentiment factor for these markets on Value-at-Risk estimation. Finally, regarding potential fat-tail and asymmetry properties between sentiment changes and marker returns, we apply new Copula-based GARCH model in our study.
