2.2. SANTA CATALINA DE SALINAS
2.2.3. DESARROLLO TURÍSTICO
To test Hypothesis 3.4, a difference-in-differences multivariate analysis of cumulative abnormal returns around SEO announcements was conducted, conditional on underwriting quality. The regressions results are displayed in Table 3.7. In particular, the sample was split based on whether an SEO issue was underwritten (Model (5)) or not (Model (6)). The underwritten SEOs were further divided based on whether the issues were underwritten by a top 10 (Model (3)) or non-top 10 (Model (4)) underwriter, or by a top 4 (Model (1)) or non-top 4 investment bank (Model (2)). The reasons for these divisions are threefold. First, underwriters can certify the quality of an issuing firm and thus relieve investors’ concerns over the negative economic impact of environmental risk (the certification hypothesis). An SEO issue was defined in this study as underwritten if the issue had data on the ‘bookrunner’ column24 extracted from the Bloomberg database. Second, the quality of the certification effect may be more highly regarded by investors if reputable investment banks underwrite the issue (the reputation hypothesis). An SEO issue is defined as underwritten by reputable investment banks
if at least one of the lead underwriters is listed on the top 10 underwriters by market share (altogether accounting for 81.6% of the market share) as recorded by Bloomberg.25 Third, the most reputable underwriters can guarantee the quality of the issuer even further, thereby probably attracting positive responses from the investors. This prediction was tested by focusing the analysis on the top 4 underwriters by market share.
This analysis was conducted only on SEOs and not on rights offers for two reasons: (1) few rights issues are underwritten, and (2) previous analysis showed weak evidence on the impact of environmental risk on rights announcement abnormal returns. This fact is understandable because the relatively low level of information asymmetry associated with rights issues disincentivises issuers from using costly underwriting services. For each condition, the public offer sample was partitioned into two subsamples accordingly and the same model regressions run on both, testing for the difference in estimated coefficients on the interaction term
POLLUTER*POST.
25 The top 10 underwriters for primary seasoned equity offerings by market share in descending order are (1) UBS,
(2) Macquarie, (3) Goldman Sachs, (4) JP Morgan, (5) Bank of America Merrill Lynch, (6) Deutsche Bank, (7) Credit Suisse, (8) Citi, (9) RBS and (10) Morgan Stanley. These were the top 10 underwriters for the Australian market over the sample period 2001–2013 as provided by Bloomberg. We also tested the hypothesis using the top 4 underwriters according to the literature and the results were similar.
Table 3.7: Environmental Risk and SEO Announcement Abnormal Returns, Conditional on Underwriting Quality
Panel A: Regression results conditional underwriting qualities
Dep. Var. 5-Day CAR
Underwriting Quality Top 4
Underwritten Non-Top 4 Underwritten
Top 10
Underwritten Non-Top 10 Underwritten
All Underwritten All Non- Underwritten Model (1) (2) (3) (4) (5) (6) Ind. Var. POLLUTER*POST 0.050* −0.016 0.014 −0.006 −0.008 −0.053** [1.80] [−1.17] [0.73] [−0.40] [−0.67] [−2.42] LNASSETS −0.001 0.001 −0.004 0.004 0.0004 −0.004 [−0.17] [0.60] [−1.06] [1.11] [0.19] [−1.27] ROA 0.055 0.024*** 0.035 0.026*** 0.026*** 0.012** [1.25] [2.86] [1.57] [2.94] [3.36] [2.07] D/E −0.002 −0.001 0.001 −0.002 −0.002 −0.004 [−0.37] [−0.39] [0.30] [−0.50] [−0.73] [−1.04] CASH −0.017 0.005 −0.013 0.007 0.001 −0.033 [−0.24] [0.30] [−0.32] [0.41] [0.09] [−1.52] TOBINQ −0.001 0.002* 0.0003 0.003** 0.002* 0.001 [−0.13] [1.69] [0.19] [2.24] [1.76] [1.10] REVOL 0.332 −0.145 −0.447 −0.073 −0.129 0.100 [0.39] [−0.92] [−0.78] [−0.44] [−0.86] [0.68] RUNUP −0.029 −0.060*** −0.038** −0.063*** −0.057*** −0.044*** [−1.11] [−8.02] [−2.12] [−7.88] [−8.11] [−6.71] Constant 0.024 0.015 0.102 −0.051 0.010 0.090 [0.14] [0.26] [1.21] [−0.77] [0.24] [1.36]
(Industry & Year) FE Yes Yes Yes Yes Yes Yes
Observations 234 1,277 417 1,094 1,511 1,103
R-squared 0.129 0.113 0.067 0.119 0.093 0.083
Panel B: P-value of two-tailed test of difference in coefficient of POLLUTER*POST
Model (1) v. (2): 0.022**
Model (3) v. (4): 0.397
Model (5) v. (6): 0.060*
Model (1) v. (6): 0.002***
Model (3) v. (6): 0.017**
Panel A of the table displays the results of the impact of polluters and the post-Kyoto period on the SEO announcement abnormal returns, conditional on the underwritten quality of the issue. The status includes issues underwritten by top 4 (Model (1), by non-top 4 (Model (2) or by top 10 (Model (3)) underwriter; by non-top 10 investment bank (Model (4)); by any investment bank (Model (5)); and not underwritten (Model (6)). Industry and Year fixed effects are included in all regressions. Robust t statistics are reported in parentheses. All the explanatory variables are defined in previous tables. Panel B of the table reports the chi-squared two-tailed test
(p-value) of the differences in the coefficients of the interaction term POLLUTER*POST across models. The *,
** and *** indicate significance at 10%, 5% and 1%, respectively.
Panel A of Table 3.7 shows that, consistent with expectations, the coefficient on the interaction term was only significantly negative on the subsample indicating non-underwritten SEOs
on the subsamples of non-top 4 (Model (2)) and non-top 10 (Model (4)) underwritten SEOs. Interestingly, the coefficients on the interactions were positive if the SEOs were underwritten by a top 10 or top 4 investment bank, but only significant for those underwritten by a top 4 underwriter.
Panel B of Table 3.7 reports the two-tailed chi-squared tests (p-value) of differences in the coefficients of the interaction term between any two models. The test results indicate that the explanatory power of the interaction term was significantly higher for SEOs that were underwritten by a top 4 than those underwritten by a non-top 4, indistinguishable between top 10 and non-top 10, and significantly lower for non-underwritten relative to underwritten SEOs. Regarding the level of the reputation of the underwriter, the effect of the interaction on cumulative abnormal returns was significantly higher for SEOs that were underwritten by either a top 4 or a top 10 over those that were non-underwritten. However, the difference was more significant at 1% for the top 4 as compared with 5% for the top 10 underwriters. In sum, the results reveal that underwriters can certify the value of SEO issuers, hence relieving investors’ concerns about the increasing information asymmetry and environmental risk associated with the issuers. And the certification value increases with the reputation of the underwriting investment banks.
3.7 ROBUSTNESS TESTS
This section presents three robustness tests. First, the event windows over which the cumulative abnormal returns were calculated were altered, and then a similar multivariate analysis was performed using the same model specifications as in the main tests. Second, a further control was implemented for the endogeneity concerns by adopting the PSM technique.