171 / 322
臺大管理論叢
第
26
卷第
3
期
171
using ARIMA model.
TTM
t
is the time-to-maturity (in days) of the highest open interest futures
contracts.
D
j
(
j
= 1, 2, 3, 4) are the four daily dummies for the day-of-the-week effect. Numbers in
parentheses for regression coefficients are t-values. The statistics of H
0
are the F-values. The
superscript a, b and c indicate significance at the 1%, 5% and 10% confidence levels,
respectively.
4.4 Test Results of Open Interest Representing Divergence in Opinions
Table 6 presents the results for the test of implications (5) to (6) in Hypothesis 3, which
relates the open interest to the divergence of opinions. In Models (7) and (8), the dependent
variables are respectively volume and the absolute value of the difference between the bid
and ask depth. The key independent variables are the signed open interest changes,
DiOI
1
|
ΔOI
| for
OI
increment and
DiOI
2
|
ΔOI
| for
OI
decrement. The control variables are
expected open interest, unexpected open interest, market depth (for Model (7) only), total
volume (for Model (8) only), time to expiration (
TTM
), and dummies for the day-of-the-
week effect (
D
j
). The regression results are summarized as follows.
Consistent with the prediction of implication (5), increments and decrements in open
interest have different impacts on the change in volume. In Model (7), one unit increment in
open interest raises volume by 5.9722 units. The impact is greater than the effect (3.4629
units) of one unit decrement in open interest on volume. The difference is statistically
significant according to the F-test for null hypothesis of
β
1
=
β
2
with an F-value of 54.42. Our
results suggest that the effect of open interest increases on trading volume is greater than the
effect of open interest decreases on trading volume. It indicates that increments of open
interest may reflect a greater degree of divergence in investors’ opinions, supporting the view
of Bessembinder et al. (1996). The positive relation between increments in open interest and
volume also is consistent with practitioners’ view that divergences in traders’ opinions will
cause volume to rise.
Model (8) examines the information contents of open interest with regard to the
symmetricity in bid and ask depths, using the absolute difference of bid-side and ask-side
depths (|
BDEP – ADEP
|) as a proxy for divergence in opinion. The larger the |
BDEP –
ADEP
| is, the more convergence the opinions are. Our regression results show a negative
β
81
.
Although the coefficient is statistically insignificant, the sign is consistent with the
conjecture that open interest increment is associated with more symmetric depths on buy-
and sell-side of the order book. The significantly positive
β
82
indicates that open interest
decrement is accompanied by a larger depth deviation between the buy- and sell-side of the
order book. Results are in general consistent with implication (6), thus support the