122 The Basics of financial economeTrics
investment-grade and noninvestment grade bonds using all observations.
We will then test to see if there is any structural break between the two
dates. We organize the data in matrix form as usual. Data are shown in
Table 6.1. The columns labeled “CCC+ and Below” indicate that data
belong to two categories and suggests the use of one dummy variable.
Another dummy variable is used later to distinguish between the two dates.
Let’s first estimate the regression equation for the fully pooled data, that
is, all data without any distinction in categories. The estimated coefficients
for the model and their corresponding t-statistics are shown as follows:
Coefficient
Estimated
CoefficientStandard
Error t-Statistic p-Value
β 0 157.01 89.56 1.753 0.081
β 1 61.27 8.03 7.630 9.98E-13
β 2 –13.20 2.27 –5.800 2.61E-08
β 3 –90.88 16.32 –5.568 8.41E-08tAble 6.1 Regression Data for the Bond Spread Application, November 28, 2005,
and June 6, 2005
Issue #
Spread,
11/28/05CCC+
and
Below CouponCoverage
RatioLogged
EBITSpread,
6/6/05CCC+
and
Below CouponCoverage
RatioLogged
EBIT
1 509 0 7.400 2.085 2.121 473 0 7.400 2.087 2.111
2 584 0 8.500 2.085 2.121 529 0 8.500 2.087 2.111
3 247 0 8.375 9.603 2.507 377 0 8.375 5.424 2.234
4 73 0 6.650 11.507 3.326 130 0 6.650 9.804 3.263
5 156 0 7.125 11.507 3.326 181 0 7.125 9.804 3.263
6 240 0 7.250 2.819 2.149 312 0 7.250 2.757 2.227
7 866 1 9.000 1.530 2.297 852 1 9.000 1.409 1.716
8 275 0 5.950 8.761 2.250 227 0 5.950 11.031 2.166
9 515 0 8.000 2.694 2.210 480 0 8.000 2.651 2.163
10 251 0 7.875 8.289 1.698 339 0 7.875 8.231 1.951
11 507 0 9.375 2.131 2.113 452 0 9.375 2.039 2.042
12 223 0 7.750 4.040 2.618 237 0 7.750 3.715 2.557
13 71 0 7.250 7.064 2.348 90 0 7.250 7.083 2.296
14 507 0 8.000 2.656 1.753 556 0 8.000 2.681 1.797
15 566 1 9.875 1.030 1.685 634 1 9.875 1.316 1.677
16 213 0 7.500 11.219 3.116 216 0 7.500 10.298 2.996
17 226 0 6.875 11.219 3.116 204 0 6.875 10.298 2.996
18 192 0 7.750 11.219 3.116 201 0 7.750 10.298 2.996