Order Imbalance Relation and Trading Strategies in NASDAQ Hedging Stocks

碩士 === 國立臺灣大學 === 財務金融學研究所 === 95 === In general, investors trade for two reasons: to hedge and share risk and to speculate on the private information. Previous research suggests that the dynamic relation between volume and returns lies in the underlying motivations. For aggressive investors, their...

Full description

Bibliographic Details
Main Authors: Chien-Chang Chiu, 邱堅彰
Other Authors: 蘇永成
Format: Others
Language:zh-TW
Published: 2007
Online Access:http://ndltd.ncl.edu.tw/handle/15560255524458085965
Description
Summary:碩士 === 國立臺灣大學 === 財務金融學研究所 === 95 === In general, investors trade for two reasons: to hedge and share risk and to speculate on the private information. Previous research suggests that the dynamic relation between volume and returns lies in the underlying motivations. For aggressive investors, their hedging actions (i.e. rotation) tend to result in abrupt price soaring and subsequent reversal in a short period of time. In this paper, by introducing specific selection criteria, we try to screen the potential targets and develop the trading strategy. The results indicate that for samples with maximum loss below 5%, the results reveal a paradox of high upside and low downside. Top three sectors account for more than half of the samples, which implies that hedge initiators seem to prefer specific sectors when screening potential rotation targets. In addition, the practice of “clearing the floats” plays a key role in analyzing the waiting period. It is found that most price jumps are likely to accompany volume augmentation, and most samples show price reversal on the jump day. Lastly, with GARCH(1,1) model, we verify the fitness of GARCH model in capturing the time variant property of returns and the relationship between order imbalance and returns. The results reveal that order imbalance indeed presents positively significant influence on returns of most samples. However, the relationship between order imbalance coefficients and market cap fails to present significance, which implies that size effect may not exist.