Table 15 - Summary of prediction errors
Estimation |
Obs. |
Prediction error (notches) |
% Correctly |
% Within 1 |
% Within 2 | ||||||||||
5 |
4 |
3 |
2 |
1 |
0 |
-1 |
-2 |
-3 |
-4 |
-5 | |||||
OLS |
557 |
0 |
5 |
~7Γ |
88 |
209 |
141 |
ɪ |
2 |
0 |
0 |
37.5% |
78.6% |
96.6% | |
RE with εi |
557 |
0 |
0 |
1 |
17 |
78 |
361 |
91 |
8 |
1 |
0 |
0 |
64.8% |
95.2% |
99.6% |
Moody’s RE without εi |
557 |
0 |
6 |
15 |
49 |
92 |
188 |
141 |
53 |
12 |
1 |
0 |
33.8% |
75.6% |
93.9% |
Ordered Probit |
557 |
4 |
4 |
14 |
35 |
99 |
259 |
86 |
46 |
10 |
0 |
0 |
46.5% |
79.7% |
94.3% |
RE Ordered Probit |
557 |
0 |
8 |
23 |
59 |
106 |
244 |
71 |
34 |
11 |
1 |
0 |
43.8% |
75.6% |
92.3% |
OLS |
568 |
0 |
3 |
15 |
34 |
104 |
218 |
147 |
41 |
6 |
0 |
0 |
38.4% |
82.6% |
95.8% |
RE with εi |
565 |
0 |
0 |
1 |
6 |
80 |
392 |
83 |
2 |
1 |
0 |
0 |
69.4% |
98.2% |
99.6% |
S&P RE without εi |
565 |
0 |
5 |
12 |
39 |
98 |
216 |
133 |
52 |
10 |
0 |
0 |
38.2% |
79.1% |
95.2% |
Ordered Probit |
565 |
0 |
10 |
14 |
28 |
99 |
262 |
118 |
23 |
10 |
1 |
0 |
46.4% |
84.8% |
93.8% |
RE Ordered Probit |
565 |
1 |
12 |
13 |
41 |
115 |
218 |
130 |
29 |
6 |
0 |
0 |
38.6% |
81.9% |
94.3% |
OLS |
481 |
1 |
3 |
6 |
32 |
87 |
196 |
113 |
43 |
0 |
0 |
0 |
40.7% |
82.3% |
97.9% |
RE with εi |
481 |
0 |
1 |
2 |
4 |
63 |
339 |
71 |
1 |
0 |
0 |
0 |
70.5% |
98.3% |
99.4% |
Fitch RE without εi |
481 |
1 |
3 |
7 |
39 |
93 |
174 |
106 |
57 |
1 |
0 |
0 |
36.2% |
77.5% |
97.5% |
Ordered Probit |
481 |
1 |
0 |
16 |
32 |
91 |
209 |
95 |
31 |
6 |
0 |
0 |
43.5% |
82.1% |
95.2% |
RE Ordered Probit |
553 |
1 |
3 |
25 |
53 |
115 |
191 |
121 |
36 |
8 |
0 |
0 |
34.5% |
77.2% |
93.3% |
Notes: * - prediction error within +/- 1 notch. ** - prediction error within +/- 2 notches.
Table 16 - Upgrades and downgrades prediction
Sample |
Predicted Upgrades |
Upgrades correctly |
Sample |
Predicted |
Downgrades correctly | |||
t |
t+1 |
t |
t+1 | |||||
OLS |
60 |
95 |
23 |
20 |
34 |
55 |
20 |
10 |
RE with εi |
60 |
87 |
28 |
17 |
34 |
51 |
16 |
12 |
Moody's RE without εi |
60 |
89 |
23 |
16 |
34 |
51 |
17 |
8 |
Ordered Probit |
60 |
127 |
31 |
25 |
34 |
72 |
20 |
8 |
RE Ordered Probit |
60 |
101 |
23 |
23 |
34 |
65 |
18 |
8 |
OLS |
79 |
79 |
32 |
17 |
41 |
50 |
16 |
15 |
RE with εi |
79 |
79 |
31 |
14 |
41 |
52 |
18 |
12 |
S&P RE without εi |
79 |
90 |
34 |
15 |
41 |
61 |
19 |
14 |
Ordered Probit |
79 |
102 |
38 |
14 |
41 |
64 |
20 |
13 |
RE Ordered Probit |
79 |
90 |
31 |
15 |
41 |
68 |
20 |
12 |
OLS |
68 |
74 |
28 |
19 |
25 |
35 |
13 |
3 |
RE with εi |
68 |
67 |
25 |
19 |
25 |
34 |
15 |
7 |
Fitch RE without εi |
68 |
89 |
24 |
20 |
25 |
53 |
15 |
5 |
Ordered Probit |
69 |
115 |
30 |
24 |
25 |
71 |
15 |
5 |
RE Ordered Probit |
89 |
154 |
43 |
29 |
26 |
77 |
13 |
7 |
Note: εi - estimated country specific effect.
49
More intriguing information
1. If our brains were simple, we would be too simple to understand them.2. The name is absent
3. The name is absent
4. Critical Race Theory and Education: Racism and antiracism in educational theory and praxis David Gillborn*
5. Linkages between research, scholarship and teaching in universities in China
6. APPLYING BIOSOLIDS: ISSUES FOR VIRGINIA AGRICULTURE
7. How do investors' expectations drive asset prices?
8. The name is absent
9. PERFORMANCE PREMISES FOR HUMAN RESOURCES FROM PUBLIC HEALTH ORGANIZATIONS IN ROMANIA
10. The name is absent