傳統財務理論中均假設市場為效率市場，然而不動產市場並非效率市場，投資者對於市場資訊之反應並非完全理性。若投資者忽視自身擁有之資訊，選擇追隨其他人的投資決策，將使投資人間存在相互牽制之行為，因而產生行為財務學中之「從眾行為」，其決策結果將無法完全反應市場資訊，並造成投資人集體買進、賣出之行為，使市場價格與交易量存在不正常之波動。由於台灣不動產市場長期以來存在有價格漲幅波動超越合理範圍之現象，因此本研究探討台灣不動產市場是否存有從眾行為，使得投資人具有非理性的投資傾向。 有鑑於過去關於從眾行為之研究仍以股票市場中報酬率或交易量驗證為主，對於台灣運用交易量進行不動產市場之從眾行為驗證則付之闕如，而從眾行為對於不動產市場之影響，首先將反映於交易量之波動，因此本研究運用自我迴歸分配落遲模型對於台灣不動產市場是否存在從眾行為進行驗證，並比較不動產報酬率波動不同之交易市場，其從眾行為存在情形之異同。 模型結果顯示台灣三大都會區（臺北市、臺中市與高雄市）與臺北市分區（分為市中心、郊區與郊外）中，僅臺北市整體與臺北市分區之住宅市場明顯存在從眾行為現象。結果顯示當該住宅市場存在從眾行為時，當期交易量將受到當期持有成本與前期市場報酬率之影響。此外，交易量除受從眾行為之影響外，尚受到經濟成長率、營建類股股價指數、物價指數租金年增率、營造工程物價指數等之正向影響，而購屋貸款利率與通貨膨脹則和交易量呈反向變動現象。 本研究以探討從眾行為、交易量與總體經濟因素之關連性，進一步釐清影響住宅市場交易量波動之因素，使購屋者於決策時參考前期市場交易情形能更加理性，避免盲目跟隨下的從眾行為產生。 / Base on the Efficient Market Hypothesis, the traditional financial theory assumes the market is efficient. However, the real estate market is not. For this reason, investors could not react to market information entirely. If investors ignore their own information, they may choose to follow other peoples’ investment decisions. Therefore, this situation will lead to herding behavior of behavioral finance that may cause price volatility and unusual transactions. On account of the real estate market exists unreasonable price fluctuations for a long time in Taiwan, this thesis examines whether the herding behavior exists in Taiwan real estate market or not. Although many researchers study the herding behavior in the stock market by using the transactions and the returns on investment, few attempts have been made to discuss the herding behavior in Taiwan housing market by using the housing transactions. Hence, this study examines the herding behavior in Taiwan housing market by establishing the Auto-regressive Distributed Lag (ARDL) model with housing transaction data. Results found the herding behavior of real estate market do exist in the whole Taipei city and the three region of Taipei city (downtown, suburb and outskirt). And it shows the transactions in the housing market with herding behavior may be affected by user cost of housing and pre-market returns. Furthermore, the study finds some macroeconomic factors affecting the housing transactions positivity, such as economic growth rate, construction stocks index, consumer price index of house renting and consumer price index of construction engineering. On the contrary, loan interest rate of housing and consumer price index has negative influence. To conclude, this study aims to examine the influential factors on the volatility of housing transactions though clarifying the relationship between the herding behavior, the transactions in housing market and the macroeconomic factors. It may help investors follow other peoples’ investment decisions more reasonable, and avoid blind herding behavior in real estate markets.

不動產由於具有異質性、不可移動性與昂貴性等特性，相較於其他資產而言，最獨特的風險為「流動性風險」(liquidity risk)，也因此其銷售期間比其他標的較長，議價空間也較大。以往探討不動產流動性的研究大都以在市場上的銷售期間(time on the market, TOM)為主，然本研究認為銷售期間和賣方的表價（listing price）與買方心中的價格（offer price）密切相關，買方決策的過程勢必受到賣方表價與本身心中價格之影響，進而決定該不動產在市場之流動性。故本研究則嘗試以賣方表價與最後成交價(transaction price)間的議價空間，探討不動產市場之流動性。議價空間愈大，表示不動產標的在當時市場之流動性愈低，致使賣方愈能接受低於表價的買方出價及成交價。 本研究透過搜尋理論建立住宅不動產議價空間模型，並以實際市場交易資料進行實證分析，探討房屋本身的屬性、總體市場因素、賣方持有的成本、區位因素與賣方定價因素等，對於買賣雙方議價空間之影響，藉以觀察理論與實證是否相符。其中以房屋總坪數與屋齡代表房屋屬性，以房租成長率與經濟成長率代表市場情況，並以利率代表賣方持有成本。實證結果顯示，屋齡太久或賣方定價過高的不動產，其議價空間愈大，流動性愈差；房租成長率和經濟成長率皆與議價空間呈負相關，表市場景氣愈好，議價空間愈小，賣方在議價過程中較能堅守其表價；又利率與議價空間呈正相關，表賣方持有不動產的成本越高，越能接受較大的議價空間；而總坪數愈大及區位較佳之不動產，其議價空間越小，可能受豪宅市場效應以及區域抗跌性有關。此實證結果與過去利用銷售期間衡量不動產市場流動性的搜尋理論相符，也驗證議價空間實可為衡量不動產市場流動性的新指標，並可降低利用銷售期間分析的研究困境。本研究成果不僅可供不動產賣方定價策略、買方議價時機之參考，亦希望透過本研究對議價空間與不動產市場流動性之研究，期望政府儘速建立與公開不動產交易平台，俾利增進不動產市場之流動性，更能牽動不動產市場與整體經濟市場之成長。 / The study suggests that not only time on market (TOM) but also price concession between the listing and contract prices could measure housing market liquidity. Departed from past studies, this paper develops theory and constructs a model named Residential Housing Price Concession Model to examine whether key factors influenced housing market liquidity significantly from past studies would have the same effect on price concession. The model includes the listing price of house, the macroeconomic data, the cost of the search and other housing characters in empirical model. Results show that listing price, cost of search and age of house have the predicted positive coefficients, and macroeconomic data , squares of house and location factor are found to be negatively related to the price concession. The corresponding conclusion with time on market (TOM) examined by past studies explains that the price concession also could measure housing market liquidity.

我國公債期貨市場發展至今，市場流動性未能有效提振，本文將針對此問題嘗試從市場結構、實務狀況、相關學理等各方面，探討諸多可能的影響因素，並加以分析研究，找出問題的癥結以提供解決之道。此外，本文亦從問券調查的結果中，歸納出市場參予者對現行公債期貨的看法，希冀能作為台灣期貨交易所未來商品規劃之參考。 / Since the Taiwan government bond futures trade, the market is lack of liquidity during the year. For the problem, this paper considers the layers of market structure, trading convention and relative theories, try to analysis the causes of less liquidity and resolve the liquidity problem. On the other hand, by making the survey this paper sums up the opinions from the participants of the bond futures market. This paper, which could help the Taiwan Futures Exchange in designing other interest rate derivatives, will wish to give some useful reference.