(可能几乎最快之）股票崩盘危机指标之指令 (III)

黃河泉 发表于 2018-12-23 07:16
1. 我的例子是用虚拟资料，所以结果不是重点 (会与一般之真正估计值有所差异)。2. 我猜你们问题最大可能是 ...

----------------------- copy starting from the next line -----------------------

1. * Example generated by -dataex-. To install: ssc install dataex
   
2. clear
   
3. input long stkcd int year str7 week float(Ri Rm)
   
4. 2 2011 "2011-31" -.032067 -.025772
   
5. 2 2011 "2011-35"  .029484  .030857
   
6. 2 2011 "2011-38" -.002475 -.004105
   
7. 2 2011 "2011-47" -.057181 -.026295
   
8. 2 2011 "2011-27"    .0358   .00928
   
9. 2 2011 "2011-15"  .037415  .019822
   
10. 2 2011 "2011-21" -.008424 -.004594
    
11. 2 2011 "2011-03"  .017998 -.021881
    
12. 2 2011 "2011-33"  .030562 -.011132
    
13. 2 2011 "2011-44"  .133238  .071492
    
14. 2 2011 "2011-07"   .00978  .015266
    
15. 2 2011 "2011-32"  .003681 -.025366
    
16. 2 2011 "2011-40" -.021622 -.035445
    
17. 2 2011 "2011-30" -.001186 -.016935
    
18. 2 2011 "2011-20"  .001205  .003674
    
19. 2 2011 "2011-16" -.027322  .003481
    
20. 2 2011 "2011-46" -.040816 -.017404
    
21. 2 2011 "2011-05"  -.01432  .012456
    
22. 2 2011 "2011-02"  .081509  .008859
    
23. 2 2011 "2011-26"  .040994  .041603
    
24. 2 2011 "2011-48"   .00141 -.013619
    
25. 2 2011 "2011-10"   .02583  .020714
    
26. 2 2011 "2011-04" -.074033 -.031556
    
27. 2 2011 "2011-11" -.001199 -.000402
    
28. 2 2011 "2011-49"  .039437 -.010821
    
29. 2 2011 "2011-09" -.026347  -.00389
    
30. 2 2011 "2011-14"  .011468 -.009988
    
31. 2 2011 "2011-12" -.015606 -.007561
    
32. 2 2011 "2011-19" -.021226 -.014598
    
33. 2 2011 "2011-39" -.081886 -.021528
    
34. 2 2011 "2011-37" -.002469 -.015663
    
35. 2 2011 "2011-22" -.029126 -.056147
    
36. 2 2011 "2011-45"  -.00885  .025347
    
37. 2 2011 "2011-08"  .010896  .027639
    
38. 2 2011 "2011-42"  .015193  .029854
    
39. 2 2011 "2011-23" -.013924  .010679
    
40. 2 2011 "2011-25"   .00625 -.019537
    
41. 2 2011 "2011-43"  -.05034 -.053235
    
42. 2 2011 "2011-06" -.009685  .015671
    
43. 2 2011 "2011-34" -.034401 -.022818
    
44. 2 2011 "2011-18" -.008187 -.037598
    
45. 2 2011 "2011-51" -.014589 -.042327
    
46. 2 2011 "2011-13"  .063415  .023042
    
47. 2 2011 "2011-50"   .02168 -.023188
    
48. 2 2011 "2011-28"  .016129  .019724
    
49. 2 2011 "2011-17" -.039326 -.010916
    
50. 2 2011 "2011-24"  .026958 -.004464
    
51. 2 2011 "2011-36" -.033413 -.032644
    
52. 2 2011 "2011-52"  .017497 -.012249
    
53. 2 2011 "2011-29" -.044218  .011783
    
54. 2 2011 "2011-53" -.011905  -.00649
    
55. 2 2012 "2012-34" -.056911 -.010224
    
56. 2 2012 "2012-21"  .031469 -.004066
    
57. 2 2012 "2012-16"  .006993  .017336
    
58. 2 2012 "2012-41" -.022539  .009566
    
59. 2 2012 "2012-52"  .073171  .040913
    
60. 2 2012 "2012-44"  .054878  .025129
    
61. 2 2012 "2012-22"  .033898  .021877
    
62. 2 2012 "2012-24"    .0213  .020354
    
63. 2 2012 "2012-12" -.018138 -.024538
    
64. 2 2012 "2012-38" -.063549 -.047183
    
65. 2 2012 "2012-28"  .021762 -.013761
    
66. 2 2012 "2012-37" -.030233 -.001358
    
67. 2 2012 "2012-20" -.039194 -.020397
    
68. 2 2012 "2012-17"  .039352 -.007403
    
69. 2 2012 "2012-15"  .023866  .024525
    
70. 2 2012 "2012-48"  .035503 -.031559
    
71. 2 2012 "2012-25" -.008782 -.016712
    
72. 2 2012 "2012-42"  .021845  .012358
    
73. 2 2012 "2012-43" -.026128  -.03239
    
74. 2 2012 "2012-36"  .072319  .045543
    
75. 2 2012 "2012-01" -.041499 -.024942
    
76. 2 2012 "2012-45" -.031214 -.027768
    
77. 2 2012 "2012-35" -.012315 -.024538
    
78. 2 2012 "2012-09"  .014101  .008667
    
79. 2 2012 "2012-13"  .019704  -.04371
    
80. 2 2012 "2012-30" -.034409  -.01988
    
81. 2 2012 "2012-51" -.027835  .004385
    
82. 2 2012 "2012-49"  .051429  .042291
    
83. 2 2012 "2012-29" -.056795 -.012783
    
84. 2 2012 "2012-31"  -.03118  .001273
    
85. 2 2012 "2012-33" -.020478 -.026797
    
86. 2 2012 "2012-05" -.013924  .007354
    
87. 2 2012 "2012-06"  .007702  .013767
    
88. 2 2012 "2012-23" -.025137 -.037361
    
89. 2 2012 "2012-46" -.019093 -.029016
    
90. 2 2012 "2012-19" -.031453 -.022227
    
91. 2 2012 "2012-14"  .012077  .024278
    
92. 2 2012 "2012-02"  .041899  .039153
    
93. 2 2012 "2012-39"  .079385  .029426
    
94. 2 2012 "2012-03"  .058981  .030807
    
95. 2 2012 "2012-32"  .010345  .022298
    
96. 2 2012 "2012-50"  .054348  .041144
    
97. 2 2012 "2012-47"  .027981  .004919
    
98. 2 2012 "2012-10" -.003476 -.004712
    
99. 2 2012 "2012-18"  .026726  .025987
    
100. 2 2012 "2012-27"  .098208  .005053
     
101. 2 2012 "2012-11" -.038372 -.013271
     
102. 2 2012 "2012-07" -.007643  .005435
     
103. 2 2012 "2012-08"  .092426  .039413
     
104. end

复制代码

415812560 发表于 2018-12-24 16:22
----------------------- copy starting from the next line -----------------------

你有没注意到你的资料顺序有问题 (week)，你怎么取落后 (lag) 或往前(forward) 项呢？

黃河泉 发表于 2018-12-23 07:16
1. 我的例子是用虚拟资料，所以结果不是重点 (会与一般之真正估计值有所差异)。2. 我猜你们问题最大可能是 ...

老师，后来我自己更改了一下代码，算出来了。跟已有文献的描述性统计也是一致的。谢谢老师~

楠瓜酥 发表于 2019-1-3 17:29
老师，后来我自己更改了一下代码，算出来了。跟已有文献的描述性统计也是一致的。谢谢老师~

Great to hear that.

黄老师您好，实在不明白您DUVOL指标的计算。我根据公式做的命令是这样的。跟您的有很大区别。您为什么要求标准差呢？希望得到老师的回复。谢谢老师了
bysort id year: egen W_mean=mean(W)
gen u=(W>W_mean)
bysort id year: egen n_u=sum(u)
gen n_d=n-n_u
bysort id year: egen W2_sum_Down=sum(W^2*(1-u))
bysort id year: egen W2_sum_Up=sum(W^2*u)
gen DUVOL=log(((n_u-1)*W2_sum_Down)/((n_d-1)*W2_sum_Up))

黃河泉 发表于 2018-12-25 10:26
你有没注意到你的资料顺序有问题 (week)，你怎么取落后 (lag) 或往前(forward) 项呢？

那我要怎么做呢？

415812560 发表于 2019-1-13 17:26
那我要怎么做呢？

试试类似

1. sort stkcd week
   
2. gen t = _n
   
3. xtset stkcd t
   
4. gen F1Rm = F1.Rm
   
5. gen F2Rm = F2.Rm
   
6. gen L1Rm = L1.Rm
   
7. gen L2Rm = L2.Rm

复制代码

copy starting from the next line        ------        ----------------

1. * Example generated by -dataex-. To install: ssc install        dataex
   
2. clear
   
3. input long id double(week1 W2 TW2)
   
4. 2 2655    .001403548232591973  .0343863288093198
   
5. 2 2656   .0007408452908540468  .0343863288093198
   
6. 2 2657   .0007671794963587059  .0343863288093198
   
7. 2 2658  .00009687958954542393  .0343863288093198
   
8. 2 2659   .0005135228307492247  .0343863288093198
   
9. 2 2660   .0006093619800145778  .0343863288093198
   
10. 2 2661 2.4429571887016552e-08  .0343863288093198
    
11. 2 2662 3.7039592985492056e-06  .0343863288093198
    
12. 2 2663  .00007367816574861226  .0343863288093198
    
13. 2 2664   .0012994262504230198  .0343863288093198
    
14. 2 2665  .00044580520780044175  .0343863288093198
    
15. 2 2666  .00016077984643596193  .0343863288093198
    
16. 2 2667    .001320490238867055  .0343863288093198
    
17. 2 2668   .0009360918478184787  .0343863288093198
    
18. 2 2669   .0011428585514634984  .0343863288093198
    
19. 2 2670 .000013640873205285516  .0343863288093198
    
20. 2 2671  .00003593307560595152  .0343863288093198
    
21. 2 2672 .000024679233685802444  .0343863288093198
    
22. 2 2673   .0011633922344138554  .0343863288093198
    
23. 2 2674    .000682424700418834  .0343863288093198
    
24. 2 2675   .0010641898283052999  .0343863288093198
    
25. 2 2676   .0008167315011331377  .0343863288093198
    
26. 2 2677  .00003733788276065515  .0343863288093198
    
27. 2 2678   .0005221860236776284  .0343863288093198
    
28. 2 2679  .00009849272318769955  .0343863288093198
    
29. 2 2680    .004103050933495359  .0343863288093198
    
30. 2 2681  .00019984618290251442  .0343863288093198
    
31. 2 2682 .000016511865157185782  .0343863288093198
    
32. 2 2683   .0009860624126788987  .0343863288093198
    
33. 2 2684   .0019435646958655712  .0343863288093198
    
34. 2 2685  .00008761238250506845  .0343863288093198
    
35. 2 2686 .000059460962477701125  .0343863288093198
    
36. 2 2687 1.9108804335920636e-06  .0343863288093198
    
37. 2 2688  .00019845035553752122  .0343863288093198
    
38. 2 2689 1.4114833737598052e-09  .0343863288093198
    
39. 2 2695    .002473543770010175  .0343863288093198
    
40. 2 2696   .0020087578486666937  .0343863288093198
    
41. 2 2697    .000688636068956483  .0343863288093198
    
42. 2 2698   .0008375284197291147  .0343863288093198
    
43. 2 2699  .00025391389195813114  .0343863288093198
    
44. 2 2700   .0022967783593564684  .0343863288093198
    
45. 2 2701   .0020612378426409844  .0343863288093198
    
46. 2 2702   .0010457309320059168  .0343863288093198
    
47. 2 2703   .0011505255995234432  .0343863288093198
    
48. 2 2704  .00010125903644969686 .03476673991097546
    
49. 2 2710   .0002200780514165509 .03476673991097546
    
50. 2 2711   .0019334405957286104 .03476673991097546
    
51. 2 2712 1.1791158576849697e-08 .03476673991097546
    
52. 2 2713  4.386896506932541e-06 .03476673991097546
    
53. 2 2714   .0007547390953194543 .03476673991097546
    
54. end
    
55. format %twCCYY-ww week1

复制代码

copy up to and including the previous        line -        ----------------

老师，想问问，按您的方法我算出来TW2是每一年的和。如何才能按公司每年每周逐项求和，让TW2等于第一周加第二周；第一周加第二周加第三周........这样

蓝君若天 发表于 2019-1-19 17:33
copy starting from the next line        ------        ----------------

copy up to and including the previous        li ...

试试

1. bys id (week1): gen wanted = sum(W2)

复制代码