While our main analysis focuses on the capital market consequences of eliminating the reconciliation requirement, we
also provide evidence on firms’ responses to the elimination. We find that none of the IFRS users continue to provide the
reconciliation in Form 20-Fs voluntarily after the elimination. Moreover, we compare the number of press releases before
and after the elimination and find that IFRS firms do not increase their disclosure frequency after the elimination.
To address the concern on whether our tests have sufficient statistical power given the relatively small sample size, we
conduct power analysis and find that our regressions do not suffer from significant Type II errors. In addition, finding
results that are consistent across different measures and robust to different model specifications increases the confidence
that our results are not caused by variable mis-measurement or model mis-specifiatoin. Specifically, we find consistent
results across four different measures of liquidity and PIN along with PIN parameters. Additional analyses of more than 10
measures of other capital market consequences also yield consistent results. To ensure a correct model specification, we
employ a difference-in-differences design, using cross-listed non-IFRS-reporting firms to control for contemporaneous
Table 7
Analysis of press releases around the elimination of the 20-F reconciliation.
Year N No. of press releases No. of press releases related to
forecasts of earnings or revenues
IFRS users (a)
2006 78 17.36 2.42
2007 78 15.10 1.85
Diff. 2.26 0.58
p-Value 0.539 0.333
Non-IFRS users (b)
2006 162 8.28 1.31
2007 162 8.07 1.19
Diff. 0.20 0.12
p-Value 0.923 0.666
a vs. b
Diff-in-diff 2.05 0.46
p-Value 0.606 0.419
Variable definitions:
Number of press releases: The number of press releases issued by the company over 90 days starting one trading day before the 20-F filing dates.
Number of press releases related to forecasts of earnings or revenues: The number of press releases related to forecasts of earnings or revenue forecasts
issued by the company over 90 days starting one trading day before the 20-F filing dates.
266 Y. Kim et al. / Journal of Accounting and Economics 53 (2012) 249–270
changes in capital markets that may coincide with the SEC’s decision to end the reconciliation requirement. Our results are
robust to alternative control samples. We also include an extensive list of control variables in our regression models to
mitigate the concern of the correlated omitted variables problem.
Taken together, we find no evidence that eliminating the IFRS to U.S. GAAP reconciliation has a negative capital market
impact on IFRS-reporting cross-listed firms. The findings of our paper will be of interest to accounting and security
regulators. Our results are inconsistent with the argument that the SEC’s decision to end the reconciliation requirement
results in information loss or greater information asymmetry. Our results suggest that U.S. investors do not perceive crosslisted
firms’ accounting information based on IFRS to be of inferior quality relative to that based on U.S. GAAP, either
because they view the two sets of standards to have similar quality, or because they believe that, holding firms’ reporting
incentives constant, differences between IFRS and U.S. GAAP have a minimal impact on the quality of accounting
information. Current and potential U.S. cross-listed firms will also find our evidence useful in their analysis of the costs and
benefits of listing their securities in the U.S. capital markets.