Anon

Regression Models with Categorical Variables 125

tAble 6.1 (continued)

Issue #

Spread,

11/28/05

CCC+

and

Below Coupon

Coverage

Ratio

Logged

EBIT

Spread,

6/6/05

CCC+

and

Below Coupon

Coverage

Ratio

Logged

EBIT

98 481 0 6.750 2.677 1.858 439 0 6.750 3.106 1.991

99 270 0 7.625 2.835 3.109 242 0 7.625 2.813 3.122

100 190 0 7.125 9.244 3.021 178 0 7.125 7.583 3.138

Notes:

Spread = option-adjusted spread (in basis points).

Coupon = coupon rate, expressed without considering percentage sign (i.e., 7.5% = 7.5).

Coverage Ratio = EBITDA divided by interest expense for company.

Logged EBIT = logarithm of earnings (EBIT in millions of dollars).

Other regression results are:

SSR: 2.3666e+006

F-statistic: 89.38

p-value: 0

R^2 : 0.57

Given the high value of the F-statistic and the p-value close to zero, the

regression is significant. The coefficient for the three regressors is statisti-

cally significant and has the expected sign. However, the intercept term is

not statistically significant. The residuals are given in the second column of

Table 6.2.

tAble 6.2 Illustration of Residuals and Leverage for Corporate Bond Spread
Issue # Residuals Residuals Dummy 1 Residuals Dummy 2
  1   118.79930   148.931400   162.198700
  2   126.39350   183.097400   200.622000
  3   –68.57770   –39.278100   –26.716500
  4   –37.26080   –60.947500   –71.034400
  5     16.63214       4.419645     –3.828890
  6 –128.76600 –104.569000   –92.122000
  7   386.42330   191.377200   217.840000
  8     73.53972     48.516800       56.58778
  9   104.15990   146.400600   160.438900
10 –124.78700   –98.020100   –71.374300
11     –4.28874     73.473220     94.555400
12 –117.58200   –88.168700   –82.883100
13 –223.61800 –213.055000 –202.748000

(continued)