DSpace university logo mark
Japanese | English 

NAOSITE : Nagasaki University's Academic Output SITE > 020 経済学部・経済学研究科 > 020 紀要 > 經營と經濟 > 第90巻 第3号 >

Empirical Exploration on the Foreign Exchange Market Microstructure with Duration, Volume and Orderflow

ファイル 記述 サイズフォーマット
keizai90_3_307.pdf749.82 kBAdobe PDF本文ファイル

タイトル: Empirical Exploration on the Foreign Exchange Market Microstructure with Duration, Volume and Orderflow
著者: Susai, Masayuki
発行日: 2010年12月24日
出版者: 長崎大学経済学会 / Economics Society of Nagasaki University
引用: 經營と經濟, vol.90(3), pp.307-327; 2010
抄録: With Ultra-High Frequent JPY against UDS rate data, we investigate the features of the market microstructure of foreign exchange rate in terms of duration, volatility, volume and orderflow. ACD model developed by Russell/Engle(1995)enables us to use tick-by-tick type foreign exchange data without any loss of information. In addition to ACD model, we use GARCH type model modified to accomplish our purpose. With ACD model, we found clear results. We confirmed that persistency of expected duration is strong. Thus, we cannot say foreign exchange market is efficient. As for the market microstructure exploralation, we use Diamond/Verrecchia(1986)model which shows that no trade means bad news, and Easly/O'Hara(1992)model which mentions that no trade means no news. Our results show that volume has significant negative impact on duration and buying USD orderflow also affects negatively on duration. If we can recognize volume and orderflow as the proxy of information inflow, foreign exchange market can be seen as the market Easly/O'Hara(1992)shown. The bigger the volume becomes, the shorter the trade interval becomes, the shorter the duration becomes, then the more the trade occurs. In the next step, we analyze the relations between volatility and duration to explore the foreign exchange market structure. Duration has negative effect on volatility, but has no effect on foreign exchange rate in 2007,whereas duration has positive effect on volatility with negative effect on USD rate in 2008.If we take the result from ACD into account, the more information comes to market and the more trade occurs, the higher volatility becomes in 2007.This result is consistent with the implication from mixture of distribution hypothesis and Easly/ O'Hara(1992)model. In 2008,the longer the trading interval becomes, the higher the volatility becomes. From ACD estimation, volume has positive impact on duration. Combining with these results, we can say that no trade, which means no news, makes market volatile. This result suggests that Diamond/Verrecchia(1986)model is valid. Uncertainty in the market was quite high and USD used to depreciate with active trade in our data in 2008.It might be said that foreign exchange market can be the market suggested by Easly/O'Hara(1992) in a usual case, but it also depends on the market condition such as uncertainty or risk.
記述: 杉原敏夫教授定年退職記念号 / In Honour of Prof. Toshio Sugihara
キーワード: market microstructure / autoregressive conditional duration model
URI: http://hdl.handle.net/10069/24917
ISSN: 02869101
資料タイプ: Departmental Bulletin Paper
原稿種類: publisher
出現コレクション:第90巻 第3号

引用URI : http://hdl.handle.net/10069/24917



Valid XHTML 1.0! Copyright © 2006-2015 長崎大学附属図書館 - お問い合わせ Powerd by DSpace