Information Asymmetry and Volume

1
Information Asymmetry and Volume Return
Relation A time varying model based on order
imbalance of individual stock

• Yongchern Su
• Hanching Huang

2
Key Issues

• Analyze the order imbalance - return effect,
  effect of return autocorrelation and volume -
  volatility effect simultaneously.
• Use intraday data ( 90 second, 15 minute ) to
  express the dynamic relation.

3
Literature review on the relation between return
and order imbalances

• Order imbalances reveal more information than
  volume.
• Large order imbalance has great impact on price
  movement.
• Signal private information (Kyle (1985))
• Exert pressure on market makers inventory and
  prompt a change in quotes (Spiegel and
  Subrahmanyam (1995))

4
Literature review on the relation between return
and order imbalances

• Order imbalances are positively autocorrelated.
• Herding (Scharfstein and Stein (1990)
• Split orders over time to reduce price impact (
  Kyle(1985))

5
Literature review on the relation between return
and order imbalances

• Chordia and Subrahmanyam (2004)
• The expectation of the price change conditional
  on the contemporaneous and lagged imbalances are
  positive and negative respectively.

6
Literature review about stock return

• At the aggregate levels , the returns tend to
  reverse themselves on high-volume days. (Campbell
  et al.(1993))
• In the individual stocks , the returns tend to
  continue on low-transactions securities.(Conrad
  et al.(1994))

7
Literature review on the relation between volume
and return

• Llorente, Michaely,Saar,and Wang (2002)(LMSW)
• reconciles the contrasting empirical results .
  Hedging trades -- returns reversalsSpeculative
  trades returns continue.

8
Literature review on the relation between volume
and volatility

• Chan and Fong (2000) Order imbalance plays a
  role in the volatility - volume relation for the
  relation becomes weaker after controlling the
  return impacts of the order imbalances.

9
Objectives

• Test the contemporaneous and lag-one order
  imbalance - return effect
• Analyze the effect of return autocorrelation and
  order imbalance on the autocorrelation of stock
  returns
• Test the volatility order imbalance effect

10
Methodology

• According to the dynamic return volume relation
  on LMSW (2002), we develop a GARCH (1,1) model.
  
• Cons
• As companies issue new shares as bonuses, stock
  price is diluted.
• Individual investors do not complain about it,
  but foreign investors believe that their holdings
  in the firm are robbed.
• Foreign government charged Taiwanese firms for
  dumping.

11
Methodology
Coefficient Effect
a1 contemporaneous order imbalance - return effect
a2 lag-one order imbalance-return effect
a3 effect of return autocorrelation
a4 effect of order imbalance on the autocorrelation of stock returns
ß3 volatility order imbalance effect
12
Data

• Data selection NASDAQ 100 stocks from Dec 2,
  2002 to Jan 6, 2003
• Select the NASDAQ 100 stocks for their market
  capitalization and trading volumes are large
  enough for insiders to trade on their private
  information.

13
NASDAQ-100 Index

• According to the dynamic return volume relation
  on LMSW (2002), we develop a GARCH (1,1) model.
  
• Cons
• As companies issue new shares as bonuses, stock
  price is diluted.
• Individual investors do not complain about it,
  but foreign investors believe that their holdings
  in the firm are robbed.
• Foreign government charged Taiwanese firms for
  dumping.

14
Descriptive statistics
15
The order imbalance return effects
16
The effect of return autocorrelation
17
The volatility order imbalance effect
18
The test between three markets within 90-second
regime
19
The test between three markets within 15-minute
regime
20
The test between 15-minute and 90-second regime
21
Conclusion

• The contemporaneous and the lag one order
  imbalance - return effects are positive and
  significant.
• The information has almost fully reflected in the
  contemporaneous price and few reflected in the
  lag-one price.

22
Conclusion

• The effect of return autocorrelation and the
  effect of order imbalance on the autocorrelation
  of stock returns are positive and significant
  within 15 - minute regime.
• The above effects are positive and insignificant
  within 90-second regime.
• The volatility order imbalance effect is
  positive and significant.

23
Conclusion

• We cannot reject the hypothesis that the effects
  are no differences between the three market
  situations within 90 second regime, but we can
  reject that within 15 minute regime.
• We can reject the hypothesis that there are no
  difference between the effects within 90 second
  regime and that within 15 minute regime.