本研究針對認定分析師預測報導對市場的影響力檢測的幾項可能缺失，經由實證設 計調整，本研究利用兩階段加權平方法，處理了虛假相關、內生性及極端值等問題，並 加入重要的控制變數提升了檢定力，有效釐清公司本身盈餘宣告及分析師盈餘預測報導 這兩個發生時間相隔極短的重要事件帶給市場的衝擊大小，”平均而言”，在公司盈餘宣 告帶給市場衝擊後，分析師盈餘預測報導確實會帶給市場另一個邊際顯著的衝擊，實證 結果介在 Altinkiliç et al.(2013)及 Bradley et al. (2014)之間，也就是說，分析師提供的盈 餘預測報導對市場反應而言是有效的，但在統計上僅具邊際顯著。
本研究以不同的訊息特性變數與公司特性變數將樣本作分類後，迴歸分析所得出的 市場反應確實會不一樣，顯示對於訊息、財報難解讀或資訊量少的公司，此時市場投資 人更需要分析師的報導。最後，在探討分析師因應速度的議題上，顯示當市場關注或較 需仰賴分析師解讀時，分析師因應速度會較快，並非僅與公司實際發佈的盈餘宣告時間 有關。
31
參考文獻
Altinkilic, O., and R. Hansen, 2009. On the information role of stock recommendation revisions. Journal of Accounting and Economics 48, 17–36.
Altinkilic, O., V. Balashov and R. Hansen, 2013, Evidence that analysts are not important information intermediaries. Management Science 59, 2550-2565.
Altinkilic, O., R. Hansen, and Y. Liyu, 2016, Can analysts pick stocks for the long-run?
Journal of Financial Economics 116, 371-398.
Asquith, P., M. Mikhail, and A. Au., 2005. Information content of equity analyst reports.
Journal of Financial Economics 75, 245-282.
Bradley, D., J. Clarke, S. Lee and C. Ornthanalai, 2014. Are analysts’ recommendations informative? Intraday evidence on the impact of time stamp delays. Journal of Finance 69, 645–673.
Barber, B., R. Lehavy, M. McNichols, and B. Trueman, 2001, Can investors profit from the prophets? Security analyst recommendations and stock returns. Journal of Finance 56, 531-563.
Edward, L., K. Ramesh, M. Shen, and J. Wu, 2015, Do analyst stock recommendations piggyback on recent corporate news? An analysis of regular-hour and after-hours revisions.
Journal of Accounting Research 53, 821- 861.
Fang, L. H., and A. Yasuda., 2014. Are stars’ opinions worth more? The relation between analyst reputation and recommendation values. Journal of Financial Services Research 46, 235-269.
Frankel, R., S. P. Kothari , and J. Weber, 2006. Determinants of the informativeness of analyst research. Journal of Accounting and Economics 41, 29–54.
Green, T. C., R. E. Jame, S. Markov, and M. Subasi, 2014, Access to management and the informativeness of analyst research.” Journal of Financial Economics 114, 239-255.
Hansen, R., 2015. What is the value of sell-side analysts? Evidence from coverage changes – A discussion. Journal of Accounting and Economics 60, 58–64.
Huang, A., A.Y. Zang, and R. Zheng, 2014. Evidence on the Information content of text in analyst reports. The Accounting Review 89(6), 2151–2180.
Ivkovic, Z., and N. Jegadeesh, 2004. The timing and value of forecast and recommendation revisions. Journal of Financial Economics 73, 433-463.
Li, K., and H. You, 2015. What is the value of sell-side analysts? Evidence from coverage initiations and terminations. Journal of Accounting and Economics 2015, 141-160.
32
Loh, R., and R. Stulz, 2011, When are analyst recommendation changes influential? Review
of Financial Studies 24, 593-627.
Matsumoto, D., Pronk, M., and Roelofsen, E. (2011). What makes conference calls useful?
The information content of managers’ presentations and analysts’ discussion sessions. The
Accounting Review 86(4), 1383–1414.
Womack, K., 1996, Do brokerage analysts’ recommendations have investment value?
Journal of Finance 51, 137–167.
33
Figure 1 公司盈餘宣告時間點(EAT)與第一位分析師報導時間點(AAT) Day 1 Day 2 N1b T1 N1a N2b T2 N2a ………
………
00:00 9:30 16:00 24:00 9:30 16:00 24:00 open close open close
盤前 盤中 盤後 盤前 盤中 盤後 (非交易) (有交易) (無交易) (非交易) (有交易) (無交易) EAT AAT EAT AAT
第 1 組(N1b,T1) 第 2 組(N1a,T2)
34
Figure 2 盈餘宣告與分析師報導對交易市場的影響
00:00 09:30 10:30 16:00 EAT open AAT close 1hr 0.5hr 1.5hr
ar1 ar2 ar3 ar4 ar5 ar6 ar7 ar8 ar9 ar10 ar11 ar12 ar13 ar14 ar15 ar16 ar17 ar18 ar19 ar20 欲研究分析師對市場影響之區間
盈餘宣告對市場主要的影響時間 區間
35
Table 1 EAT 與 AAT 在同一日曆天
(同一天)分析師發佈盈餘預測報導時點(AAT)
盤前 盤中 盤後
公司發佈實際盈餘數 字時點(EAT)
盤 前
40 (2.6%)
634 (41.14%)
348 (22.58%) 盤
中
69 (4.48%)
81 (5.26%) 盤
後
369 (23.95%)
36
Table 2 EAT 與 AAT 相距一日曆天
(隔一天)分析師發佈盈餘預測報導時點(AAT)
盤前 盤中 盤後
公司發佈實際盈餘數 字時點(EAT)
盤 前
185 (15.34%)
81 (6.72%)
17 (1.41%) 盤
中
77 (6.38%)
42 (3.48%)
13 (1.08%) 盤
後
487 (40.38%)
222 (18.41%)
82 (6.8%)
37
Table 3 法說會時間
Panel A 法說會日與盈餘宣告日
公司數(家) 百分比(%)
在盈餘宣告日後的隔天 7 0.30
在盈餘宣告日後的二天以上 86 3.75
在盈餘宣告日前的二天以上 15 0.65
在盈餘宣告日的當天，且晚於分析師報導 76 3.31
在盈餘宣告日的當天，且早於分析師報導 1,982 86.32
在盈餘宣告日的當天，但無法確定是晚於或早於分析師報導 110 4.79
沒有法說會 20 0.87
總和 2296 100.00
Panel B 盈餘宣告日當天，盈餘宣告、法說會、分析師報導發生順序
發生順序(早至晚) 相距時間 公司數(家) 平均值(秒)
法說會、盈餘宣告、分析師報導 法說會時點 - 盈餘宣告時點
分析師報導時點 - 法說會時點
24 24
-2037.5000 15097.5000
法說會=盈餘宣告、分析師報導 法說會時點 - 盈餘宣告時點
分析師報導時點 - 法說會時點
8 8
0.0000 18367.5000
盈餘宣告、分析師報導、法說會 法說會時點 - 盈餘宣告時點
分析師報導時點 - 法說會時點
76 76
24715.2632 -7485.7895
盈餘宣告、法說會、分析師報導 法說會時點 - 盈餘宣告時點
分析師報導時點 - 法說會時點
1,950 1,950
8786.7533 12691.7867
Panel C 盈餘宣告日當天，法說會過後的分析師報導
公司數(家) 百分比(%)
一小時內 64 3.23
一小時以上~二小時內 156 7.87
二小時以上~三小時內 420 21.19
三小時以上~四小時內 605 30.52
四小時以上~五小時內 443 22.35
五小時以上~六小時內 259 13.07
六小時以上 35 1.77
總和 1,982 100.00
38
Table 4 樣本剔除情形
公司數(家) 百分比(%)
情境一 277 12.06
情境二 555 24.17
情境三 1,120 48.78
39
40
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept
+/- 0.0001027 (2.46)** 0.0000727 (1.79)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0608400 (-34.25)** -0.0719685 (-2.27)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0608600 (-34.28)** -0.0724550 (-2.27)WLS (dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
41
Table 6 (continued)
Panel D
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0001061 (-3.65)** -0.0000726 (-1.95)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0001061 (-3.65)** -0.0000729 (-1.95)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0001061 (-3.65)** -0.0000731 (-1.95)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0001048 (-1.09) -0.0000728 (-0.94)
iT+/-
0.0004117 (1.45)+ 0.0006305 (2.57)*N
6,635 8Adj. R-sq.
0.000242
Table 6 (continued)
Panel H
iT iT
iT
iT
＝
0
1
2* T
a
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0001027 (-1.07) -0.0000724 (-0.94)
iT+
-0.0004634 (-0.66) -0.0000111 (-0.02)T
a iT*
-
0.0000423 (1.36)+ 0.0000320 (0.78)N
6,635 8Adj. R-sq.
0.0003Panel I
iT iT
iT
＝
0
1
T=11~13 [分析師報導後的 30 分鐘]
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0004002 (-2.48)** -0.0004492 (-2.76)*
iT+
0.0011500 (2.45)** 0.0012673 (2.35)*N
2,279 8Adj. R-sq.
0.0022
T=14~20 [分析師報導後的 30~90 分鐘]
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0006817 (-2.73)** -0.0006849 (-2.08)*
iT+
0.0018200 (2.51)** 0.0017891 (2.90)*N
2,263 8Adj. R-sq.
0.0023Panel J
iT iT
iT
＝
0
1
T=11~13 [分析師報導後的 30 分鐘]
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0004001 (-2.48)** -0.0004492 (-2.76)*
iT+
0.0011500 (2.44)** 0.0012698 (2.36)*N
2,279 8Adj. R-sq.
0.0022
T=14~20 [分析師報導後的 30~90 分鐘]
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0006815 (-2.73)** -0.0006835 (-2.08)*
iT+
0.0018200 (2.51)** 0.0017842 (2.92)*N
2,263 8Adj. R-sq.
0.0023✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
43
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0000922 (-2.27)* -0.0000694 (-1.59)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0000932 (-2.3)* -0.0000689 (-1.60)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0001983 (-0.88) -0.0002220 (-0.98)Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0006994 (-2.02)* -0.0008311 (-3.83)**✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 7 假說 1b 分析師預測修正方向之資訊意涵 WLS(dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
44
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0001061 (-3.69)** -0.0000712 (-2.01)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0001072 (-3.69)** -0.0000709 (-1.99)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0004053 (-2.51)** -0.0004616 (-2.84)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0006859 (-2.75)** -0.0007223 (-2.05)*✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 8 假說 2a 年未預期盈餘與分析師報導時市場反應的關係 WLS(dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
45
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0000968 (-3.15)** -0.0000660 (-1.55)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0000968 (-3.15)** -0.0000656 (-1.54)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0004349 (-2.54)** -0.0004647 (-2.94)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0005620 (-2.13)* -0.0005616 (-1.46)+TEMP
=1，暫時性盈餘大於中位數，否則為 0；此次的暫時性盈餘=SPI+FCA+XI+SPPIV；✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 9 假說 2b 暫時性盈餘與分析師報導時市場反應的關係 WLS(dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
46
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0001083 (-3.72)** -0.0000780 (-2.09)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0001082 (-3.72)** -0.0000780 (-2.08)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0004023 (-2.49)** -0.0004444 (-2.79)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0007024 (-2.82)** -0.0006850 (-2.21)*✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 10 假說 2c 盈餘波動與分析師報導時市場反應的關係 WLS(dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
47
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0000995 (-3.29)** -0.0000653 (-1.63)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0000994 (-3.29)** -0.0000650 (-1.60)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0004394 (-2.61)** -0.0004790 (-3.12)**Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0005816 (-2.23)* -0.0005558 (-1.63)+✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 11 假說 3a 小公司與分析師報導時市場反應的關係 WLS(dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
48
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0001043 (-3.59)** -0.0000702 (-1.84)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0001042 (-3.58)** -0.0000702 (-1.82)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0003962 (-2.45)** -0.0004365 (-2.63)*Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0006630 (-2.66)** -0.0006621 (-1.95)*✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 12 假說 3b 分析師追蹤人數與分析師報導時市場反應的關係 WLS(dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
49
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0001086 (-3.11)** -0.0000743 (-1.49)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0001086 (-3.11)** -0.0000736 (-1.45)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.000420 (-2.15)* -0.0004275 (-3.38)**Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0007297 (-2.43)** -0.0007275 (-1.68)+✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 13 假說 3c 裁量性應計數與分析師報導時市場反應的關係 WLS(dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
50
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0000957 (-2.89)** -0.0000656 (-1.45)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0000957 (-2.89)** -0.0000668 (-1.47)+Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0003926 (-2.13)* -0.0004276 (-3.32)**Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.0006346 (-2.82)** -0.0005737 (-1.35)✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 14 假說 3d 實質盈餘操控與分析師報導時市場反應的關係 WLS(dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
51
Panel A 公司盈餘宣告在盤中、分析師盈餘報導在公司宣告時點的 24 小時內 應變數：1，分析師因應時間短(或快)
應變數：0，分析師因應時間長(或慢)
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-1.1446000 (52.94)** -1.2600182 (-5.75)**AbsAbCohen +/-
0.2446000 (4.55)* 0.2148268 (1.16)AbsUE_meanRatiostd +/-
0.0159000 (0.50) 0.1003494 (1.70)+TEMP2ratio +/-
-0.0280000 (2.55)+ -0.0511212 (-1.25)Intanratio +/-
0.0383000 (0.04) -0.0963947 (-0.34)Stdeps_7y +/-
-0.0086800 (0.19) 0.0186942 (0.49)LogAT +/-
0.1152000 (25.99)** 0.1241400 (3.55)**NUM_ANA +/-
0.0252000 (26.93)** 0.0292525 (3.46)**N
1,423 8Panel B
公司盈餘宣告在盤後、分析師盈餘報導在公司宣告時點的 24 小時內應變數：1，分析師因應時間短(或快) 應變數：0，分析師因應時間長(或慢)
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/-
-0.1056000 (3.66)* -0.0793623 (-0.62)AbsAbCohen +/-
0.1211000 (9.94)** 0.1156889 (4.14)**AbsUE_meanRatiostd +/-
-0.0262000 (5.36)* -0.0219085 (-0.72)TEMP2ratio +/-
-0.0068000 (0.89) -0.0213717 (-2.84)*Intanratio +/-
0.1733000 (8.34)** 0.1776184 (2.72)*Stdeps_7y +/-
-0.0151000 (6.91)** -0.0221435 (-2.02)*LogAT +/-
-0.0248000 (9.29)** -0.0280070 (-1.47)+NUM_ANA +/-
0.0197000 (152.17)** 0.0206048 (11.18)**N
12,037 8✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 15 假說 4a 與 4b 分析師何時會趕工- 波比模型(probit model)
52
Panel A 公司盈餘宣告在盤中、分析師盈餘報導在公司宣告時點的 24 小時內 應變數：分析師因應時間
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
AbsAbCohen +/-
0.2493400 (8.77)** 0.2708561 (4.01)**AbsUE_meanRatiostd +/-
-0.0297300 (2.45)+ -0.0506392 (-1.15)TEMP2ratio +/-
0.0114100 (0.65) 0.0074958 (0.37)Intanratio +/-
0.4310300 (8.84)** 0.4287216 (1.64)+Stdeps_7y +/-
-0.0155600 (1.42) -0.0228493 (-1.69)+LogAT +/-
-0.0587100 (-10.46)** -0.0596265 (-3.84)**NUM_ANA +/-
0.0402400 (140.52)** 0.0470156 (7.52)**N
1,747 8Panel B
公司盈餘宣告在盤後、分析師盈餘報導在公司宣告時點的 24 小時內應變數：分析師因應時間
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
AbsAbCohen +/-
0.1832400 (36.89)** 0.1846425 (6.28)**AbsUE_meanRatiostd +/-
-0.0349200 (-14.74)** -0.0549238 (-2.84)*TEMP2ratio +/-
-0.0070100 (1.38) -0.0141429 (-1.97)*Intanratio +/-
0.4357100 (84.91)** 0.4512366 (6.74)**Stdeps_7y +/-
-0.0208800 (-23.57)** -0.0241524 (-2.46)*LogAT +/-
-0.0427500 (-43.06)** -0.0425298 (-2.94)*NUM_ANA +/-
0.0350100 (860.14)** 0.0355817 (26.98)**N
14,417 8✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 16 假說 4a 與 4b 分析師何時會趕工- Cox regression
53
Panel A 公司盈餘宣告在盤中、分析師盈餘報導在公司宣告時點的 24 小時內 應變數：1，分析師先報導
應變數：0，分析師後報導
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.1987000 (0.73) -0.2583528 (-0.80)AbsAbCohen +/-
0.3081000 (2.49)+ 0.4121152 (1.09)AbsUE_meanRatiostd +/-
0.0493000 (0.88) -0.0884541 (-0.22)TEMP2ratio +/-
0.0620000 (4.18)* 0.1634217 (0.99)Intanratio +/-
0.0285000 (0.01) -0.1148448 (-0.39)Stdeps_7y +/-
0.0202000 (0.34) 0.0624976 (1.12)LogAT +/-
0.0036100 (0.02) 0.0115643 (0.32)NUM_ANA +/-
-0.0006500 (0.01) -0.0001199 (-0.03)N
1,423 8✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table17 假說 5a 與 5b 盈餘宣告時間相同，先報導誰-波比模型(probit model)
54
Spearman 相關係數, N = 1,225
Prob > |r| (位於 H0 底下): Rho=0
NUM_FIRMno NUM_ANAno NUM_FIRMno 1.00000 0.22093
<0.0001
NUM_ANAno 0.22093
<0.0001
1.00000
NUM_FIRMno：公司盈餘宣告的順序；NUM_ANAno：分析師報導的順序。
✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 18 斯皮爾曼等級相關(Spearman rank correlation)
55
Table 19 假說 1a 分析師報導之市場資訊意涵-敏感性測試(控制法說會) WLS (dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
Panel D
iT iT
iT
iT
D D
＝
0
1*
11_13
2*
14_20
(A)剔掉法說會落在分析師盈餘預測修正報導之前 1.5 小時~之後 1.5 小時的樣本
(B)剔掉法說會落在分析師盈餘預測修正報導之前 2.5 小時~之後 1.5 小時的樣本
(C)剔掉法說會落在分析師盈餘預測修正報導之前 3.5 小時~之後 1.5 小時的樣本
Full Sample Fama-MacBeth Full Sample Fama-MacBeth Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Intercept +/- -0.0001138 (-3.67)** -0.0000809 (-2.01)* -0.0001084 (-3.25)** -0.0000764 (-1.94)* -0.0000932 (-2.35)** -0.0000586 (-1.17)
11_13 iT
* D
+ 0.0004112 (2.51)** 0.0003894 (2.25)* 0.0005191 (2.96)** 0.0005141 (2.75)* 0.0002677 (1.29)✝ 0.0001314 (0.48)14_20 iT
* D
+ 0.0003093 (2.85)** 0.0003276 (4.21)** 0.0003473 (2.98)** 0.0003857 (3.67)** 0.0002910 (2.10)* 0.0002815 (2.26)*N 19,594 8 16,904 8 11,413 8
Adj. R-sq. 0.0006 0.0009 0.0004
Panel E
iT iT
iT iT
iT
iT
D D D D
＝
0
1*
11_13
2*
14_20
3* T
a*
11_13
4* T
a*
14_20
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Intercept +/- -0.0001138 (-3.67)** -0.0000810 (-2.01)* -0.0001083 (-3.25)** -0.0000767 (-1.94)* -0.0000932 (-2.35)** -0.0000588 (-1.17)
11_13 iT
* D
+ 0.0005766 (2.23)* 0.0009364 (3.22)** 0.0005786 (2.08)* 0.0010239 (2.96)* 0.0003370 (1.02) 0.0006200 (1.85)*14_20 iT
* D
+ 0.0008363 (2.44)** 0.0009516 (2.05)* 0.0007265 (1.98)* 0.0008732 (1.99)* 0.0006537 (1.50)✝ 0.0007504 (1.42)✝11_13 a
* T
*
D
iT - -0.0000165 (-0.83) -0.0000549 (-2.67)* -0.0000059 (-0.28) -0.0000513 (-1.93)* -0.0000069 (-0.27) -0.0000490 (-1.82)✝14_20 a
* T
* D
iT - -0.0000089 (-1.62)✝ -0.0000105 (-1.44)✝ -0.0000064 (-1.09) -0.0000082 (-1.10) -0.0000061 (-0.88) -0.0000080 (-0.94)N 19,594 8 16,904 8 11,413 8
Adj. R-sq. 0.0007 0.0009 0.0003
56
Table 19 (continued)
Panel F
iT iT
iT iT
iT
iT
D D D D
＝
0
1*
11_13
2*
14_20
3* T
a*
11_13
4* T
a*
14_20
(A)剔掉法說會落在分析師盈餘預測修正報導之前 1.5 小時~之後 1.5 小時的樣本
(B)剔掉法說會落在分析師盈餘預測修正報導之前 2.5 小時~之後 1.5 小時的樣本
(C)剔掉法說會落在分析師盈餘預測修正報導之前 3.5 小時~之後 1.5 小時的樣本
Full Sample Fama-MacBeth Full Sample Fama-MacBeth Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Intercept +/- -0.0001138 (-3.67)** -0.0000812 (-2.01)* -0.0001083 (-3.25)** -0.0000768 (-1.93)* -0.0000932 (-2.35)** -0.0000594 (-1.18)
11_13 iT
* D
+ 0.0005798 (2.24)* 0.0009127 (3.18)** 0.0005779 (2.08)* 0.0009805 (2.85)* 0.0003383 (1.03) 0.0005790 (1.70)✝14_20 iT
* D
+ 0.0008492 (2.48)** 0.0010243 (2.39)* 0.0007361 (2.00)* 0.0009428 (2.22)* 0.0006617 (1.52)✝ 0.0007855 (1.51)✝11_13 a
* T
* D
iT - -0.0000171 (-0.85) -0.0000567 (-2.77)* -0.0000062 (-0.29) -0.0000514 (-1.96)* -0.0000073 (-0.29) -0.0000505 (-1.88)*14_20 a
* T
* D
iT - -0.0000091 (-1.66)* -0.0000117 (-1.71)✝ -0.0000065 (-1.11) -0.0000093 (-1.26) -0.0000062 (-0.89) -0.0000086 (-1.02)N 19,594 8 16,904 8 11,413 8
Adj. R-sq. 0.0007 0.0009 0.0003
Panel I
iT iT
iT
＝
0
1
T=11~13 [分析師報導後的 30 分鐘]
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Intercept +/- -0.0003697 (-2.17)* -0.0004079 (-3.01)** -0.0002431 (-1.34)✝ -0.0002861 (-1.73)✝ -0.0002295 (-1.10) -0.0002919 (-1.02)*
iT + 0.0012800 (2.60)** 0.0016091 (3.52)** 0.0014900 (2.84)** 0.0019503 (3.82)** 0.0007203 (1.20) 0.0008740 (1.03)*N 2,003 8 1,728 8 1,168 8
57
Adj. R-sq. 0.0029 0.0041 0.0004
T=14~20 [分析師報導後的 30~90 分鐘]
(A)剔掉法說會落在分析師盈餘預測修正報導之前 1.5 小時~之後 1.5 小時的樣本
(B)剔掉法說會落在分析師盈餘預測修正報導之前 2.5 小時~之後 1.5 小時的樣本
(C)剔掉法說會落在分析師盈餘預測修正報導之前 3.5 小時~之後 1.5 小時的樣本
Full Sample Fama-MacBeth Full Sample Fama-MacBeth Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Intercept +/- -0.0008282 (-3.11)** -0.0008229 (-2.20)* -0.0008896 (-3.12)** -0.0009018 (-1.92)* -0.0006935 (-2.03)* -0.0006892 (-1.64)✝
iT + 0.0020300 (2.63)** 0.0017493 (3.14)** 0.0022200 (2.69)** 0.0021426 (2.82)* 0.0017900 (1.82)* 0.0014105 (1.89)*N 1,990 8 1,716 8 1,159 8
Adj. R-sq. 0.0030 0.0036 0.0020
Table 19 (continued) Panel J
iT iT
iT
＝
0
1
T=11~13 [分析師報導後的 30 分鐘]
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Intercept +/- -0.0003698 (-2.17)* -0.0004084 (-3.01)** -0.0002431 (-1.34)✝ -0.0002858 (-1.72)✝ -0.0002295 (-1.10) -0.0002909 (-1.02)
iT + 0.0012800 (2.59)** 0.0016156 (3.52)** 0.0014900 (2.83)** 0.0019517 (3.79)** 0.0007176 (1.20) 0.0008711 (1.03)N 2,003 8 1,728 8 1,168 8
Adj. R-sq. 0.0029 0.0040 0.0004
T=14~20 [分析師報導後的 30~90 分鐘]
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Coefficient t-statistic Mean of
Coefficient t-statistic Intercept +/- -0.0008282 (-3.11)** -0.0008220 (-2.20)* -0.0008895 (-3.12)** -0.0009006 (-1.92)* -0.0006936 (-2.03)* -0.0006871 (-1.63)✝
iT + 0.0020300 (2.63)** 0.0017461 (3.16)** 0.0022100 (2.68)** 0.0021395 (2.82)* 0.0017900 (1.82)* 0.0014071 (1.88)*N 1,990 8 1,716 8 1,159 8
Adj. R-sq. 0.0030 0.0036 0.0020
Panel A、B、C 與 Table 6 的 Panel A、B、C 相同，故此處略。
✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
58
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/- -0.0001050 (-3.55)** -0.0000789 (-2.41)*
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/- -0.0001049 (-3.54)** -0.0000778 (-2.35)*
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/- -0.0004221 (-2.56)** -0.0004671 (-2.31)*
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/- -0.0006595 (-2.59)** -0.0006524 (-2.19)*
✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 20 假說 1b 分析師預測修正方向之資訊意涵假說-敏感性測試(REV 與 UE 方向一致時)
WLS (dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
59
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/- -0.0000887 (-3.02)** -0.0000638 (-1.78)✝
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/- -0.0000887 (-3.02)** -0.0000641 (-1.79)✝
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/- -0.0003380 (-2.07)* -0.0003819 (-2.37)*
REV
i + 0.0011200 (2.67)** 0.0014115 (2.90)**N 2,279 8
Adj. R-sq. 0.0027
T=14~20 [分析師報導後的 30~90 分鐘]
Variable Predict. Coefficient t-statistic Mean of
Coefficient t-statistic
Intercept +/- -0.0005803 (-2.30)* -0.0005856 (-1.78)*
REV
i + 0.0018200 (2.81)** 0.0014062 (2.27)*N 2,263 8
Adj. R-sq. 0.0031
Panel A 與 Table 6 的 Panel A 相同，故此處略。
✝, * and ** denote the two-tailed significance level of 10%, 5%, and 1%, respectively.
Table 21 假說 1a 分析師報導之市場資訊意涵-對比性測試(兩階段迴歸分析的另 一簡化作法，但未控制住市況)
WLS (dealing with outliers and influential points)
EAT 落在當天盤前(00:00~09:30)，AAT 落在當天 09:30~16:00
60
Full Sample Fama-MacBeth
Variable Predict. Coefficient t-statistic Mean ofCoefficient t-statistic
Intercept +/-
-0.0001111 (-3.52)** -0.0000811 (-1.91)✝Variable Predict. Coefficient t-statistic Mean of
Variable Predict. Coefficient t-statistic Mean of