由市場微觀結構論探討台灣10年期公債期貨日內不對稱的價量關係

本文運用bi- LAN STAR模式以市場微觀結捨論探討台灣10年期公債期貨日內不對稱的價量關係，擷取資料資料包含日內每15分錯與30分鐘、以及每1小時分時交易量變動與報酬波動率，依據研究目的成立四個研究命題，針對開盤、收盤量與整體交易期間進行實證，包含：（1）繪製日內隨著開盤至收盤經過的價量變數是否呈現U型。（2）判斷價量之間存在的量先價行關條。（3）受到移轉函數為非線性且具有統計顯著性的影響，價量具有非對稱性與非線性關係。（4）當交易量變動越高、則報酬波動率變化也越大，價量呈現不對稱性關係。交易量變動與報酬波動率之間的關聯性實證結果如下：（1）日內隨著開盤至收盤經過的價量圖型不完全為U型而為U+W型。交易量變動與報酬波動率隨著日內時間經過呈現開盤與收盤階段較高而其它時段較低的現象，而開盤時段相較於整體市場表現更具統計上顯著的低交易量與報酬，因此，開盤分時發生逆選擇將比其它時段機會高，且因流動性偏低不利於機構法人進行操作。（2）由於公債期貨的交易不夠頻繁且過於集中，因此，在進行模式估計時，以每1小時擷取的開盤資料較具以上命題所述的特質，呈現量先價行，而受到轉換函數為非線性所影響，價量亦具有非線性關係，受非線性式影響的不對稱性轉換強度為0.6938。而在開盤交易量變動較大時，對報酬波動率的影響不僅具有顯著性且係數值也大於1為2.3781，也就是交易量變動對報酬波動率的影響力較大，價量具有不對稱且非性性關係。This paper applies bi-ANSTAR model to investigate on Taiwan 10-year Government Bond Futures. With concept of market microstructure theory, we research asymmetrical price-volume relationship between volume change and return volatility. The study data are measured by intraday intervals of 15 minutes, 30 minutes and 1 hour. According research purposes, we establish four research propositions to study and compare the interval between first open, last close and total period. The propositions major on: (a) Draw volume change and return volatility from open to close and adjudge the grape is U sharp or not. (b) Adjudge the lead relationship between price and volume. (c) Study nonlinear transition function effect on nonlinear. price-volume relationship. (d) Compare the effect between volume change and return volatility. It seems exist large change in volume usually make magnitude effect on return volatility. Price-volume relationship is asymmetrical. Two major findings obtain regarding the price-volume relationship : (a) The exchange time intervals from open to close, the graphs of volume change and return volatility is not U sharp, but they're U+W sharp . In open and close period, volume change and return volatility is higher than other intraday periods. Especially compare with total exchange period, exchange volume and return are significant lower in open interval. It means open period has higher opportunity occur adverse selection and lower market liquidity. That makes institution investors reduce ask or bid in open interval. (b) Becausc exchanges aren't frequently and concentrate on any exchange period, only I hour intervals data present significant price-volume relationship. Volume change leads return volatility one interval, that's 1hour. Because nonlinear transition function effect, price-volume relationship is nonlinear The nonlinear and asymmetrical effect on transition strength is 0.6938. Compare with higher volume change in open period, the effect coefficient of volume change on return volatility is more than I, that's 2.3781. Volume change would effect return volatility more magnitude. Price-volume relationship is asymmetrical and nonlinear.