Table 3 Underlying mechanism.

From: Does the supply of tax information affect financial restatements? Evidence from the launch of Taxation Administration Information System III in China

 

Restatement

 

Stock price synchronicity

Forecast error

Bid-ask spread

 

SynchronicH

SynchronicL

FerrorH

FerrorL

SpreadH

SpreadL

Variables

(1)

(2)

(3)

(4)

(5)

(6)

Panel A. Stock price informativeness

CTAIS

−0.052***

−0.006

−0.028*

−0.026

−0.044**

−0.019

 

(−3.54)

(−0.35)

(−1.85)

(−1.60)

(−2.20)

(−1.40)

Firm Cons

Yes

Yes

Yes

Yes

Yes

Yes

Prov Cons

Yes

Yes

Yes

Yes

Yes

Yes

Firm Fix

Yes

Yes

Yes

Yes

Yes

Yes

Year Fix

Yes

Yes

Yes

Yes

Yes

Yes

Obs

13,506

11,731

14,165

11,231

9349

15,949

R-squared

0.469

0.494

0.470

0.478

0.511

0.448

 

Restatement

 

Institutional Holding

Online Forum

Company to Bank Distance

 

InstitutionL

InstitutionH

OnlineForumL

OnlineForumH

DistanceH

DistanceL

Variables

(1)

(2)

(3)

(4)

(5)

(6)

Panel B. Outside monitoring

CTAIS

−0.034**

−0.023

−0.056***

−0.027

−0.035**

−0.025*

 

(−2.22)

(−1.61)

(−3.85)

(−1.63)

(−2.50)

(−1.72)

Firm Cons

Yes

Yes

Yes

Yes

Yes

Yes

Prov Cons

Yes

Yes

Yes

Yes

Yes

Yes

Firm Fix

Yes

Yes

Yes

Yes

Yes

Yes

Year Fix

Yes

Yes

Yes

Yes

Yes

Yes

Observations

12,963

12,968

12,735

12,497

13,114

13,108

R-squared

0.457

0.419

0.479

0.455

0.419

0.415

  1. Source: Data for Institutional Holdings is calculated based on data from the CSMAR database. Data for the online forum is calculated based on data from the CNRDS database. Data on the distance from company to bank is from the China Banking and Insurance Regulatory Commission https://xkz.cbirc.gov.cn/jr/.
  2. This table presents the heterogeneous effects of information asymmetry on our main results. The sample used is the same as that in Table 2. Panel A uses stock price informativeness to represent information asymmetry. Columns (1) and (2) of Panel A partition the sample by stock price synchronicity (Piotroski and Roulstone, 2004), while columns (3) and (4) break out the sample by analysts’ forecast error (Eames and Glover, 2003). Columns (5) to (6) of Panel A divide the sample by bid-ask-spread (Roll, 1984). Panel B uses outside monitoring as a proxy of information asymmetry. Columns (1) and (2) of Panel B partition the sample by institutional holdings, measured by the shares owned by institutional investors. Columns (3) and (4) of Panel B break out the sample by the number of online stock forums representing individual investor attention. Columns (5) and (6) of Panel B divide the sample by the distance from the company to the nearest bank, reflecting the proximity of the bank. All regressions include firm-level and province-level controls, as defined in Table 2. The t-statistics presented in parentheses are computed using standard errors clustered by firm. All regressions incorporate Firm and Year-fixed effects. Levels of statistical significance are indicated as follows: *** for 1%, ** for 5%, and * for 10%.
Back to article page