Sample selection and distribution of initial senior bonds
We retrieve all issued senior bonds between 2005 and 2020 by non-financial public firms located in the same countries as the hybrid bond issuing firms in a first step. This results in 9,646 issued bonds by 1,684 public non-financial firms. In a next step, we exclude bonds with missing data for the coupon and issued amount. Lastly, we exclude bonds that are issued by firms that do not report in accordance with IFRSs. Overall, our initial senior bond sample from which we derive the sample for our determinants analysis (H1) and announcement effect analysis (H2) includes 6,749 senior bonds issued by 1,030 firms on 5,208 announce days.
Table B.1: Initial ample selection and distribution of senior bonds
This table reports the selection steps of the initial sample selection for our senior bond sample in Panel A. Panel B provides an overview of the distribution of the number of issuances and the issued volume of hybrid bonds by year, industry, and country in our sample. Volume denotes issue volume in billion USD.
Panel A: Sample selection
# Step Firms
Issued bonds
Announce days 1 All senior bonds issued by public non-financial firms between
2005 and 2020 (Source: Refinitiv Bond Screener) located in the same countries as hybrid issuing firms.
1,684 9,646 7,441
2 Less bonds with missing data (Coupon, Issued amount) 1,567 9,264 7,174 3 Less bonds issued by non-IFRS reporting firms 1,030 6,749 5,208
= Initial senior bond sample 1,030 6,749 5,208
Panel B: Sample distribution
Capital issued by year Capital issued by industry
Year Issues Volume Industry Issues Volume Share
2005 118 49.07 Applied Resources 143 31 1.10%
2006 144 64.40 Automobiles & Auto Parts 289 144 5.19%
2007 142 76.70 Chemicals 258 114 4.10%
2008 126 68.21 Consumer Goods and Services 623 207 7.45%
2009 265 162.87 Energy 693 419 15.10%
2010 280 131.12 Food, Beverages and Drug Retailing 523 261 9.40%
2011 298 125.41 Healthcare Services & Equipment 106 32 1.15%
2012 493 220.06 Industrial & Commercial Services 397 136 4.91%
2013 473 189.96 Industrial Goods 353 146 5.25%
2014 514 185.30 Mineral Resources 388 206 7.43%
2015 408 152.86 Real Estate 1,270 228 8.23%
2016 427 209.84 Retailers 99 34 1.21%
2017 734 264.09 Software, and IT 250 102 3.67%
2018 735 253.14 Telecommunications Services 680 423 15.22%
2019 802 264.41 Transportation 253 101 3.63%
2020 790 359.03 Utilities 424 194 6.97%
Total 6,749 2,776 Total 6,749 2,776 100%
Table B.1 (continued)
Capital issued by country
Country Issues Volume Share Country Issues Volume Share
AT 96 29.33 1.06% IT 301 166.18 5.99%
AU 122 40.51 1.46% JP 136 111.53 4.02%
BE 134 62.93 2.27% KR 20 0.64 0.02%
BM 24 9.28 0.33% LU 171 85.61 3.08%
CA 784 343.79 12.38% MN 0 0.00 0.00%
CH 224 54.83 1.97% MT 17 1.07 0.04%
CL 0 0.00 0.00% MX 286 157.37 5.67%
CN 10 3.05 0.11% MY 30 2.08 0.07%
DE 593 287.38 10.35% NL 209 147.83 5.32%
DK 57 28.70 1.03% NO 368 88.91 3.20%
ES 131 47.15 1.70% NZ 34 2.71 0.10%
FI 134 42.03 1.51% OM 0 0.00 0.00%
FR 1,380 675.91 24.34% PH 2 0.85 0.03%
GB 455 256.78 9.25% PT 39 9.84 0.35%
HK 11 1.41 0.05% SE 771 91.60 3.30%
IN 3 0.20 0.01% SG 207 26.96 0.97%
Total 4,158 1,883 67.83% Total 6,749 2,776 100.00%
Sample selection of determinants sample (H1)
We start the selection process for the sample to test H1 (determinants sample) with all issued bonds identified in our initial senior bond sample. We proceed to delete observations with missing values for the analyzed determinants. After deleting bond-year observations with missing values for lagged Size, lagged MTB, and lagged InterestCov, we arrive at our final sample of 6,007 senior bonds issued by 918 firms and 383 hybrid bonds issued by 167 firms.
Table B.2: Sample selection of determinants sample (H1)
This table reports the selection steps of the sample selection for the determinants sample to test H1.
Senior bond sample
Hybrid bond sample
# Step Firms Bonds Firms Bonds
1 Initial bond sample 1,030 6,749 179 421
2 Drop obs. with missing values for lagged Size 1,013 6,643 178 417 3 Drop obs. with missing values for lagged MTB 958 6,338 172 404 4 Drop obs. with missing values for lagged InterestCov 918 6,007 167 383 5 Final sample to test H1 (determinants sample) 918 6,007 167 383
Sample selection of announcement effects sample (H2)
To test announcement effects around senior bond, hybrid bond and seasoned equity offering announcements, we collect senior bond offering announcements and seasoned equity offering announcements of firms that also issue hybrid bonds in a first step. We therefore limit our analysis of announcement effects to selected hybrid bond sample firms. We then remove observations with missing values for CAR resulting in 328 hybrid bond announcements, 1,304 senior bond announcements, and 646 seasoned equity announcements. In a last step, we remove observations with missing values for control variables resulting in 312 hybrid bond announcements, 1,240 senior bond announcements, and 622 seasoned equity announcements of 174 sample firms.
Table B.3: Sample selection of announcement effects sample (H2)
This table reports the selection process of the sample selection for the announcement effects sample to test H2.
Announce days
# Step
Hybrid bonds
Senior bonds
Seasoned equity 1 Offering announcements of hybrid bond sample firms
from 2005 to 2020 (initial hybrid bond sample)
332 1,369 656
2 Less observations with missing cumulative abnormal returns around announcement days
328 1,304 646
3 Less observations with missing values Size, Leverage, ROA, Tangibility, and MTB
312 1,240 622
4 Final sample to test H2 (announcement effects sample) 312 1,240 622
Sample selection of value relevance sample (H3)
To test the value relevance of hybrid bonds’ book values, we hand-collect financial statement data on hybrid bonds in a first step. For 175 firms of our 179 sample firms, we were able to verify and hand-collect the hybrid bonds’ book values resulting in a total of 950 firm-year observations. We then eliminate firm-year observations with missing price data, negative equity and missing per share data resulting in a sample of 838 firm-year observations.
Table B.4: Sample selection of value relevance sample (H3)
This table reports the selection process of the sample selection for the announcement effects sample to test H2.
# Step
Firms Firm- years
Firms Firm- years 1 Firm-year observations with non-missing values
for
book values of hybrid bonds
131 716 44 234
2 Less firm-year observations with missing price data
130 707 44 232
3 Less firm-year observations with negative equity 128 660 41 203 4 Less firm-year observations with per share data 124 635 41 203
4 124 635 41 203
Table B.5 Descriptive statistics
This table reports summary statistics of firms that issue hybrid bonds classified as equity or debt and sample mean differences. All variables are defined in Appendix A. ***, **, and * denote significant differences at the 1%, 5%, and 10% levels, respectively.
Panel A: Determinants sample by type of hybrid bond issuer
Debt hybrid bond issuer Equity hybrid bond issuer Difference Variable N Mean S.D. P50 N Mean S.D. P50 Diff. t-stat Size 115 24.252 1.284 24.539 268 23.654 1.812 23.550 -0.598*** 3.67 ROA 115 0.059 0.046 0.056 268 0.048 0.038 0.041 -0.011** 2.32 SDROA 115 0.039 0.025 0.034 268 0.051 0.061 0.035 0.012*** 2.62 Leverage 115 0.286 0.128 0.239 268 0.347 0.120 0.332 0.061*** 4.36 Tangibility 115 0.346 0.227 0.354 268 0.284 0.197 0.276 -0.061** 2.52 InterestCov 115 15.923 47.729 4.786 268 11.884 61.920 3.595 -4.038 0.69 MTB 115 2.009 1.745 1.724 268 1.440 1.783 1.102 -0.569*** 2.91 Divpayer 115 0.948 0.223 1.000 268 0.910 0.286 1.000 -0.037 1.38 Cash ETR 115 0.119 0.627 0.191 268 0.919 6.904 0.201 0.800* 1.88 Credit Rating 115 0.661 0.475 1.000 268 0.466 0.500 0.000 -0.194*** 3.62 Watchlist 115 0.035 0.184 0.000 268 0.007 0.086 0.000 -0.027 1.52 AnalystFol 115 18.122 10.349 17.000 268 17.056 10.836 18.000 -1.066 0.91 Covenants 115 0.000 0.000 0.000 268 0.027 0.183 0.000 0.027** 2.39 Bonus Pay 43 0.404 0.216 0.430 102 0.409 0.223 0.430 0.006 0.15
Table B.6
Determinants of hybrid bond issuances and equity classification of hybrid bonds
This table presents the regression results from logit regressions. In columns (1) and (2), we report results for the determinants of firms’ decision to issue hybrid or senior bonds. The dependent variable Hybrid is an indicator variable that takes value one (zero) if the firm issues a hybrid bond (senior bond). In columns (3) and (4), we report results for the determinants of firms’ decision to issue hybrid bonds classified as equity or debt under IAS 32. The dependent variable EqClass is an indicator variable that takes value one (zero) if the firm issues a hybrid bond classified as equity (debt). All other variables are defined in Appendix A. All continuous variables are standardized to have zero mean and a standard deviation of one. We report robust standard errors in parentheses. ***, **, and * denote significant differences at the 1%, 5%, and 10% levels, respectively.
Hybrid vs. senior bonds Equity vs. debt classification Exp.
sign
Hybrid Hybrid Exp.
sign
EqClass EqClass
(1) (2) (3) (4)
Non-reporting incentives
Size ? 0.101 0.246 ? -0.417* -0.731
(0.098) (0.199) (0.252) (0.562)
ROA - -0.226*** -0.082 ? -0.264 -0.838
(0.061) (0.145) (0.230) (0.836)
SDROA + -0.175*** -0.071 + 0.317 1.110
(0.067) (0.067) (0.266) (1.217)
Leverage + 0.043 -0.100 + 0.416** 0.537
(0.077) (0.156) (0.175) (0.536)
Tangibility - 0.014 0.191 ? -0.218 -0.762
(0.090) (0.150) (0.216) (0.720)
InterestCov + 0.052* -0.196 + 0.046 -2.352
(0.031) (0.700) (0.132) (1.589)
MTB + -0.009 -0.147 ? -0.139 -0.053
(0.018) (0.229) (0.202) (0.677)
Divpayer ? 0.476** -0.236 ? -0.874 1.382
(0.240) (0.508) (0.731) (3.343)
Cash ETR + 0.072* 0.011 ? 3.497** 1.585
(0.038) (0.025) (1.494) (2.896)
Reporting incentives
Credit Rating + 1.841*** 2.165*** ? -0.571 -1.990
(0.158) (0.269) (0.415) (1.441)
Watchlist - -0.957** -1.022 ? -0.656
(0.466) (0.834) (0.765)
AnalystFol ? -0.183* -0.376** ? 0.844*** 1.753***
(0.100) (0.181) (0.237) (0.475)
Covenants ? -0.109 +
(0.093)
Bonus Pay ? 0.104 + 0.817**
(0.154) (0.321)
Observations 6,390 2,229 368 122
Adjusted R² 0.168 0.226 0.305 0.484
Industry-FE Yes Yes Yes Yes
Year-FE Yes Yes Yes Yes
Table B.7
Announcement effects of hybrid bond issuances
This table presents (differences in) cumulative abnormal returns (CAR) around a three-day window of offering announcements for seasoned equity offerings (Equity, Hybrid = 0), straight debt offerings (Debt, Hybrid = 0), equity- labelled hybrid bonds (Equity, Hybrid = 1) and debt-labelled hybrid bonds (Debt, Hybrid = 1). Robust t-statistics are reported in brackets. Panel A reports CARs for announcements with information on the hybrid bond’s accounting classification. Panel B reports CARs for announcements without information on the hybrid bond’s accounting classification.
Panel A: Announcement with information on hybrid bond’s accounting classification Hybrid = 0
(N=1,950)
Hybrid = 1 (N=63)
Difference (N=2,013) Equity
(N=703)
-0.0334***
[11.655]
-0.0008 [0.326]
-0.0326***
[3.359]
Debt (N=1,310)
0.0042**
[2.056]
-0.0186*
[2.112]
0.0229 [0.757]
Difference (N=2,013)
-0.0377***
[10.632]
0.0178**
[2.053]
0.0555***
[6.030]
Panel B: Announcement without information on hybrid bond’s accounting classification Hybrid = 0
(N=1,950)
Hybrid = 1 (N=171)
Difference (N=2,121) Equity
(N=757)
-0.0334***
[11.655]
-0.0006 [0.197]
-0.0328***
[4.662]
Debt (N=1,364)
0.0042**
[2.056]
0.0047 [1.526]
-0.0005 [0.053]
Difference (N=2,121)
-0.0377***
[10.632]
-0.0053 [1.118]
0.0323***
[5.780]
Table B.8
Value relevance of hybrid bonds (different outlier correction)
This table presents regression results from estimating specification (5) using stock price three months after fiscal year end as dependent variable using different techniques to accommodate for outliers. Column 1 reports results without outlier correction. Column 2 reports results from estimating robust regressions. Column 3 reports results when excluding studentized residuals larger than two. Column 4 reports results when winsorizing each variable at the 1st and 99th percentile. All variables are defined in Appendix A. We report robust standard errors clustered at the firm- level in parentheses. ***, **, and * denote significant differences at the 1%, 5%, and 10% levels, respectively.
Panel A: Value relevance of equity-labelled hybrid bonds
Price Price Price Price
(1) (2) (3) (4)
BVEPS ex HBPS 0.553*** 0.587*** 0.614*** 0.520***
(0.074) (0.006) (0.024) (0.076)
HBPS 1.179*** 0.944*** 0.965*** 1.049***
(0.353) (0.029) (0.107) (0.288)
NIPS 1.709*** 1.983*** 1.777*** 2.301***
(0.578) (0.044) (0.241) (0.558)
Observations 635 634 601 635
R-squared 0.829 0.989 0.930 0.850
Country-fixed effects Yes Yes Yes Yes
Year-fixed effects Yes Yes Yes Yes
Outlier correction None robust regression studentized res. winsorized Panel B: Value relevance of debt-labelled hybrid bonds
Price Price Price Price
(1) (2) (3) (4)
TAPS 1.590*** 2.403*** 2.018*** 1.507***
(0.517) (0.078) (0.354) (0.538)
TLPS ex HBPS -1.699*** -2.520*** -2.147*** -1.590***
(0.569) (0.088) (0.389) (0.577)
HBPS -0.581 -2.163*** -0.776 -0.510
(2.636) (0.371) (1.719) (3.188)
NIPS 2.012* -1.059*** 0.960 2.411**
(0.998) (0.243) (0.692) (1.126)
Observations 203 203 189 203
Adjusted R² 0.772 0.959 0.856 0.776
Country-fixed effects Yes Yes Yes Yes
Year-fixed effects Yes Yes Yes Yes
Outlier correction None robust regression studentized res. winsorized