-
Notifications
You must be signed in to change notification settings - Fork 15
/
fisher_rsrs.py
1344 lines (1091 loc) · 51.1 KB
/
fisher_rsrs.py
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
74
75
76
77
78
79
80
81
82
83
84
85
86
87
88
89
90
91
92
93
94
95
96
97
98
99
100
101
102
103
104
105
106
107
108
109
110
111
112
113
114
115
116
117
118
119
120
121
122
123
124
125
126
127
128
129
130
131
132
133
134
135
136
137
138
139
140
141
142
143
144
145
146
147
148
149
150
151
152
153
154
155
156
157
158
159
160
161
162
163
164
165
166
167
168
169
170
171
172
173
174
175
176
177
178
179
180
181
182
183
184
185
186
187
188
189
190
191
192
193
194
195
196
197
198
199
200
201
202
203
204
205
206
207
208
209
210
211
212
213
214
215
216
217
218
219
220
221
222
223
224
225
226
227
228
229
230
231
232
233
234
235
236
237
238
239
240
241
242
243
244
245
246
247
248
249
250
251
252
253
254
255
256
257
258
259
260
261
262
263
264
265
266
267
268
269
270
271
272
273
274
275
276
277
278
279
280
281
282
283
284
285
286
287
288
289
290
291
292
293
294
295
296
297
298
299
300
301
302
303
304
305
306
307
308
309
310
311
312
313
314
315
316
317
318
319
320
321
322
323
324
325
326
327
328
329
330
331
332
333
334
335
336
337
338
339
340
341
342
343
344
345
346
347
348
349
350
351
352
353
354
355
356
357
358
359
360
361
362
363
364
365
366
367
368
369
370
371
372
373
374
375
376
377
378
379
380
381
382
383
384
385
386
387
388
389
390
391
392
393
394
395
396
397
398
399
400
401
402
403
404
405
406
407
408
409
410
411
412
413
414
415
416
417
418
419
420
421
422
423
424
425
426
427
428
429
430
431
432
433
434
435
436
437
438
439
440
441
442
443
444
445
446
447
448
449
450
451
452
453
454
455
456
457
458
459
460
461
462
463
464
465
466
467
468
469
470
471
472
473
474
475
476
477
478
479
480
481
482
483
484
485
486
487
488
489
490
491
492
493
494
495
496
497
498
499
500
501
502
503
504
505
506
507
508
509
510
511
512
513
514
515
516
517
518
519
520
521
522
523
524
525
526
527
528
529
530
531
532
533
534
535
536
537
538
539
540
541
542
543
544
545
546
547
548
549
550
551
552
553
554
555
556
557
558
559
560
561
562
563
564
565
566
567
568
569
570
571
572
573
574
575
576
577
578
579
580
581
582
583
584
585
586
587
588
589
590
591
592
593
594
595
596
597
598
599
600
601
602
603
604
605
606
607
608
609
610
611
612
613
614
615
616
617
618
619
620
621
622
623
624
625
626
627
628
629
630
631
632
633
634
635
636
637
638
639
640
641
642
643
644
645
646
647
648
649
650
651
652
653
654
655
656
657
658
659
660
661
662
663
664
665
666
667
668
669
670
671
672
673
674
675
676
677
678
679
680
681
682
683
684
685
686
687
688
689
690
691
692
693
694
695
696
697
698
699
700
701
702
703
704
705
706
707
708
709
710
711
712
713
714
715
716
717
718
719
720
721
722
723
724
725
726
727
728
729
730
731
732
733
734
735
736
737
738
739
740
741
742
743
744
745
746
747
748
749
750
751
752
753
754
755
756
757
758
759
760
761
762
763
764
765
766
767
768
769
770
771
772
773
774
775
776
777
778
779
780
781
782
783
784
785
786
787
788
789
790
791
792
793
794
795
796
797
798
799
800
801
802
803
804
805
806
807
808
809
810
811
812
813
814
815
816
817
818
819
820
821
822
823
824
825
826
827
828
829
830
831
832
833
834
835
836
837
838
839
840
841
842
843
844
845
846
847
848
849
850
851
852
853
854
855
856
857
858
859
860
861
862
863
864
865
866
867
868
869
870
871
872
873
874
875
876
877
878
879
880
881
882
883
884
885
886
887
888
889
890
891
892
893
894
895
896
897
898
899
900
901
902
903
904
905
906
907
908
909
910
911
912
913
914
915
916
917
918
919
920
921
922
923
924
925
926
927
928
929
930
931
932
933
934
935
936
937
938
939
940
941
942
943
944
945
946
947
948
949
950
951
952
953
954
955
956
957
958
959
960
961
962
963
964
965
966
967
968
969
970
971
972
973
974
975
976
977
978
979
980
981
982
983
984
985
986
987
988
989
990
991
992
993
994
995
996
997
998
999
1000
# author:linbirg
# 2019-09-04
# 本身存在的问题：
# 1 fisher选股因子是长期，适合尽量久拿
# 2.fisher中低估值（估计）存在不稳定性
# 3.rsrs择时存在亏损过大的情况
# 优化点：1 增加止损 2 市场整体有风险时，平掉亏损仓位。 3 有风险时，除了亏损，rsrs为卖的也平掉。
import pandas as pd
import datetime as dt
class FileHelper:
@staticmethod
def dict_to_file(dict_data,path='history_rsrs.json'):
write_file(path, str(dict_data))
@staticmethod
def load_dict_from(path='history_rsrs.json'):
import json
try:
data = read_file(path)
str_data = str(data,'utf-8')
str_data = str_data.replace("'", '"')
dict_data= json.loads(str_data)
return dict_data
except Exception as e:
log.info('加载文件失败。'+str(e))
class BzUtil():
# 去极值
@staticmethod
def fun_winsorize(rs, type, num):
# rs为Series化的数据
rs = rs.dropna().copy()
low_line, up_line = 0, 0
if type == 1: # 标准差去极值
mean = rs.mean()
#取极值
mad = num*rs.std()
up_line = mean + mad
low_line = mean - mad
elif type == 2: #中位值去极值
rs = rs.replace([-np.inf, np.inf], np.nan)
median = rs.median()
md = abs(rs - median).median()
mad = md * num * 1.4826
up_line = median + mad
low_line = median - mad
elif type == 3: #Boxplot 去极值
if len(rs) < 2:
return rs
mc = sm.stats.stattools.medcouple(rs)
rs.sort()
q1 = rs[int(0.25*len(rs))]
q3 = rs[int(0.75*len(rs))]
iqr = q3-q1
if mc >= 0:
low_line = q1-1.5*np.exp(-3.5*mc)*iqr
up_line = q3+1.5*np.exp(4*mc)*iqr
else:
low_line = q1-1.5*np.exp(-4*mc)*iqr
up_line = q3+1.5*np.exp(3.5*mc)*iqr
rs[rs < low_line] = low_line
rs[rs > up_line] = up_line
return rs
#标准化
@staticmethod
def fun_standardize(s,type):
'''
s为Series数据
type为标准化类型:1 MinMax,2 Standard,3 maxabs
'''
data=s.dropna().copy()
if int(type)==1:
rs = (data - data.min())/(data.max() - data.min())
elif type==2:
rs = (data - data.mean())/data.std()
elif type==3:
rs = data/10**np.ceil(np.log10(data.abs().max()))
return rs
#中性化
@staticmethod
def fun_neutralize(s, df, module='pe_ratio', industry_type=None, level=2, statsDate=None):
'''
参数：
s为stock代码 如'000002.XSHE' 可为list,可为str
moduel:中性化的指标 默认为PE
industry_type:行业类型(可选), 如果行业不指定，全市场中性化
返回：
中性化后的Series index为股票代码 value为中性化后的值
'''
s = df[df.code.isin(list(s))]
s = s.reset_index(drop = True)
s = pd.Series(s[module].values, index=s['code'])
s = BzUtil.fun_winsorize(s,1,3)
if industry_type:
stocks = BzUtil.fun_get_industry_stocks(industry=industry_type, level=level, statsDate=statsDate)
else:
stocks = list(get_all_securities(['stock'], date=statsDate).index)
df = df[df.code.isin(stocks)]
df = df.reset_index(drop = True)
df = pd.Series(df[module].values, index=df['code'])
df = BzUtil.fun_winsorize(df,1, 3)
rs = (s - df.mean())/df.std()
return rs
@staticmethod
def unpaused(stock_list):
current_data = get_current_data()
tmpList = []
for stock in stock_list:
if not current_data[stock].paused:
# or stock in positions_list:
tmpList.append(stock)
return tmpList
@staticmethod
def remove_st(stock_list, statsDate):
current_data = get_current_data()
return [s for s in stock_list if not current_data[s].is_st]
@staticmethod
def remove_limit_up(stock_list):
h = history(1, '1m', 'close', stock_list, df=False, skip_paused=False, fq='pre')
h2 = history(1, '1m', 'high_limit', stock_list, df=False, skip_paused=False, fq='pre')
tmpList = []
for stock in stock_list:
if h[stock][0] < h2[stock][0]:
tmpList.append(stock)
return tmpList
# 剔除上市时间较短的产品
@staticmethod
def fun_delNewShare(current_dt, equity, deltaday):
deltaDate = current_dt.date() - dt.timedelta(deltaday)
tmpList = []
for stock in equity:
if get_security_info(stock).start_date < deltaDate:
tmpList.append(stock)
return tmpList
# 行业列表
@staticmethod
def fun_get_industry(cycle=None):
# cycle 的参数：None取所有行业，True取周期性行业，False取非周期性行业
industry_dict = {
'A01':False,# 农业 1993-09-17
'A02':False,# 林业 1996-12-06
'A03':False,# 畜牧业 1997-06-11
'A04':False,# 渔业 1993-05-07
'A05':False,# 农、林、牧、渔服务业 1997-05-30
'B06':True, # 煤炭开采和洗选业 1994-01-06
'B07':True, # 石油和天然气开采业 1996-06-28
'B08':True, # 黑色金属矿采选业 1997-07-08
'B09':True, # 有色金属矿采选业 1996-03-20
'B11':True, # 开采辅助活动 2002-02-05
'C13':False, # 农副食品加工业 1993-12-15
'C14':False,# 食品制造业 1994-08-18
'C15':False,# 酒、饮料和精制茶制造业 1992-10-12
'C17':True,# 纺织业 1992-06-16
'C18':True,# 纺织服装、服饰业 1993-12-31
'C19':True,# 皮革、毛皮、羽毛及其制品和制鞋业 1994-04-04
'C20':False,# 木材加工及木、竹、藤、棕、草制品业 2005-05-10
'C21':False,# 家具制造业 1996-04-25
'C22':False,# 造纸及纸制品业 1993-03-12
'C23':False,# 印刷和记录媒介复制业 1994-02-24
'C24':False,# 文教、工美、体育和娱乐用品制造业 2007-01-10
'C25':True, # 石油加工、炼焦及核燃料加工业 1993-10-25
'C26':True, # 化学原料及化学制品制造业 1990-12-19
'C27':False,# 医药制造业 1993-06-29
'C28':True, # 化学纤维制造业 1993-07-28
'C29':True, # 橡胶和塑料制品业 1992-08-28
'C30':True, # 非金属矿物制品业 1992-02-28
'C31':True, # 黑色金属冶炼及压延加工业 1994-01-06
'C32':True, # 有色金属冶炼和压延加工业 1996-02-15
'C33':True, # 金属制品业 1993-11-30
'C34':True, # 通用设备制造业 1992-03-27
'C35':True, # 专用设备制造业 1992-07-01
'C36':True, # 汽车制造业 1992-07-24
'C37':True, # 铁路、船舶、航空航天和其它运输设备制造业 1992-03-31
'C38':True, # 电气机械及器材制造业 1990-12-19
'C39':False,# 计算机、通信和其他电子设备制造业 1990-12-19
'C40':False,# 仪器仪表制造业 1993-09-17
'C41':True, # 其他制造业 1992-08-14
'C42':False,# 废弃资源综合利用业 2012-10-26
'D44':True, # 电力、热力生产和供应业 1993-04-16
'D45':False,# 燃气生产和供应业 2000-12-11
'D46':False,# 水的生产和供应业 1994-02-24
'E47':True, # 房屋建筑业 1993-04-29
'E48':True, # 土木工程建筑业 1994-01-28
'E50':True, # 建筑装饰和其他建筑业 1997-05-22
'F51':False,# 批发业 1992-05-06
'F52':False,# 零售业 1992-09-02
'G53':True, # 铁路运输业 1998-05-11
'G54':True, # 道路运输业 1991-01-14
'G55':True, # 水上运输业 1993-11-19
'G56':True, # 航空运输业 1997-11-05
'G58':True, # 装卸搬运和运输代理业 1993-05-05
'G59':False,# 仓储业 1996-06-14
'H61':False,# 住宿业 1993-11-18
'H62':False,# 餐饮业 1997-04-30
'I63':False,# 电信、广播电视和卫星传输服务 1992-12-02
'I64':False,# 互联网和相关服务 1992-05-07
'I65':False,# 软件和信息技术服务业 1992-08-20
'J66':True, # 货币金融服务 1991-04-03
'J67':True, # 资本市场服务 1994-01-10
'J68':True, # 保险业 2007-01-09
'J69':True, # 其他金融业 2012-10-26
'K70':True, # 房地产业 1992-01-13
'L71':False,# 租赁业 1997-01-30
'L72':False,# 商务服务业 1996-08-29
'M73':False,# 研究和试验发展 2012-10-26
'M74':True, # 专业技术服务业 2007-02-15
'N77':False,# 生态保护和环境治理业 2012-10-26
'N78':False,# 公共设施管理业 1992-08-07
'P82':False,# 教育 2012-10-26
'Q83':False,# 卫生 2007-02-05
'R85':False,# 新闻和出版业 1992-12-08
'R86':False,# 广播、电视、电影和影视录音制作业 1994-02-24
'R87':False,# 文化艺术业 2012-10-26
'S90':False,# 综合 1990-12-10
}
industry_list = []
if cycle == True:
for industry in list(industry_dict.keys()):
if industry_dict[industry] == True:
industry_list.append(industry)
elif cycle == False:
for industry in list(industry_dict.keys()):
if industry_dict[industry] == False:
industry_list.append(industry)
else:
industry_list = list(industry_dict.keys())
return industry_list
# 一级行业列表
@staticmethod
def fun_get_industry_levelI(industry=None):
industry_dict = {
'A':['A01', 'A02', 'A03', 'A04', 'A05'] #农、林、牧、渔业
,'B':['B06', 'B07', 'B08', 'B09', 'B11'] #采矿业
,'C':['C13', 'C14', 'C15', 'C17', 'C18', 'C19', 'C20', 'C21', 'C22', 'C23', 'C24', 'C25', 'C26', 'C27', 'C28', 'C29', 'C30', 'C31', 'C32',\
'C33', 'C34', 'C35', 'C36', 'C37', 'C38', 'C39', 'C40', 'C41', 'C42'] #制造业
,'D':['D44', 'D45', 'D46'] #电力、热力、燃气及水生产和供应业
,'E':['E47', 'E48', 'E50'] #建筑业
,'F':['F51', 'F52'] #批发和零售业
,'G':['G53', 'G54', 'G55', 'G56', 'G58', 'G59'] #交通运输、仓储和邮政业
,'H':['H61', 'H62'] #住宿和餐饮业
,'I':['I63', 'I64', 'I65'] #信息传输、软件和信息技术服务业
,'J':['J66', 'J67', 'J68', 'J69'] #金融业
,'K':['K70'] #房地产业
,'L':['L71', 'L72'] #租赁和商务服务业
,'M':['M73', 'M74'] #科学研究和技术服务业
,'N':['N78'] #水利、环境和公共设施管理业
#,'O':[] #居民服务、修理和其他服务业
,'P':['P82'] #教育
,'Q':['Q83'] #卫生和社会工作
,'R':['R85', 'R86', 'R87'] #文化、体育和娱乐业
,'S':['S90'] #综合
}
if industry:
return industry_dict[industry]
return industry_dict
# 根据行业取股票列表
@staticmethod
def fun_get_industry_stocks(industry, level=2, statsDate=None):
if level == 2:
stock_list = get_industry_stocks(industry, statsDate)
elif level == 1:
industry_list = BzUtil.fun_get_industry_levelI(industry)
stock_list = []
for industry_code in industry_list:
tmpList = get_industry_stocks(industry_code, statsDate)
stock_list = stock_list + tmpList
stock_list = list(set(stock_list))
else:
stock_list = []
return stock_list
@staticmethod
def filter_without(stocks, bad_stocks):
tmpList = []
for stock in stocks:
if stock not in bad_stocks:
tmpList.append(stock)
return tmpList
@staticmethod
def filter_intersection(stocks,others):
ret = list(set(stocks) & set(others))
return ret
class QuantLib():
@classmethod
def get_fundamentals_sum(cls, table_name=indicator, search=indicator.adjusted_profit, statsDate=None):
# 取最近的五个季度财报的日期
def __get_quarter(table_name, statsDate):
'''
返回最近 n 个财报的日期
返回每个股票最近一个财报的日期
'''
# 取最新一季度的统计日期
if table_name == 'indicator':
q = query(indicator.code, indicator.statDate)
elif table_name == 'income':
q = query(income.code, income.statDate)
elif table_name == 'cash_flow':
q = query(cash_flow.code, cash_flow.statDate)
elif table_name == 'balance':
q = query(balance.code, balance.statDate)
df = get_fundamentals(q, date = statsDate)
stock_last_statDate = {}
tmpDict = df.to_dict()
for i in range(len(list(tmpDict['statDate'].keys()))):
# 取得每个股票的代码，以及最新的财报发布日
stock_last_statDate[tmpDict['code'][i]] = tmpDict['statDate'][i]
df = df.sort_values("statDate",ascending= False)
# 取得最新的财报日期
last_statDate = df.iloc[0,1]
this_year = int(str(last_statDate)[0:4])
this_month = str(last_statDate)[5:7]
if this_month == '12':
last_quarter = str(this_year) + 'q4'
last_two_quarter = str(this_year) + 'q3'
last_three_quarter = str(this_year) + 'q2'
last_four_quarter = str(this_year) + 'q1'
last_five_quarter = str(this_year - 1) + 'q4'
elif this_month == '09':
last_quarter = str(this_year) + 'q3'
last_two_quarter = str(this_year) + 'q2'
last_three_quarter = str(this_year) + 'q1'
last_four_quarter = str(this_year - 1) + 'q4'
last_five_quarter = str(this_year - 1) + 'q3'
elif this_month == '06':
last_quarter = str(this_year) + 'q2'
last_two_quarter = str(this_year) + 'q1'
last_three_quarter = str(this_year - 1) + 'q4'
last_four_quarter = str(this_year - 1) + 'q3'
last_five_quarter = str(this_year - 1) + 'q2'
else: #this_month == '03':
last_quarter = str(this_year) + 'q1'
last_two_quarter = str(this_year - 1) + 'q4'
last_three_quarter = str(this_year - 1) + 'q3'
last_four_quarter = str(this_year - 1) + 'q2'
last_five_quarter = str(this_year - 1) + 'q1'
return last_quarter, last_two_quarter, last_three_quarter, last_four_quarter, last_five_quarter, stock_last_statDate
# 查财报，返回指定值
def __get_fundamentals_value(table_name, search, myDate):
'''
输入查询日期
返回指定的财务数据，格式 dict
'''
if table_name == 'indicator':
q = query(indicator.code, search, indicator.statDate)
elif table_name == 'income':
q = query(income.code, search, income.statDate)
elif table_name == 'cash_flow':
q = query(cash_flow.code, search, cash_flow.statDate)
elif table_name == 'balance':
q = query(balance.code, search, balance.statDate)
df = get_fundamentals(q, statDate = myDate).fillna(value=0)
tmpDict = df.to_dict()
stock_dict = {}
name = str(search).split('.')[-1]
for i in range(len(list(tmpDict['statDate'].keys()))):
tmpList = []
tmpList.append(tmpDict['statDate'][i])
tmpList.append(tmpDict[name][i])
stock_dict[tmpDict['code'][i]] = tmpList
return stock_dict
# 得到最近 n 个季度的统计时间
last_quarter, last_two_quarter, last_three_quarter, last_four_quarter, last_five_quarter, stock_last_statDate = __get_quarter(table_name, statsDate)
last_quarter_dict = __get_fundamentals_value(table_name, search, last_quarter)
last_two_quarter_dict = __get_fundamentals_value(table_name, search, last_two_quarter)
last_three_quarter_dict = __get_fundamentals_value(table_name, search, last_three_quarter)
last_four_quarter_dict = __get_fundamentals_value(table_name, search, last_four_quarter)
last_five_quarter_dict = __get_fundamentals_value(table_name, search, last_five_quarter)
tmp_list = []
stock_list = list(stock_last_statDate.keys())
for stock in stock_list:
tmp_dict = {}
tmp_dict['code'] = stock
value_list = []
if stock in last_quarter_dict:
if stock_last_statDate[stock] == last_quarter_dict[stock][0]:
value_list.append(last_quarter_dict[stock][1])
if stock in last_two_quarter_dict:
value_list.append(last_two_quarter_dict[stock][1])
if stock in last_three_quarter_dict:
value_list.append(last_three_quarter_dict[stock][1])
if stock in last_four_quarter_dict:
value_list.append(last_four_quarter_dict[stock][1])
if stock in last_five_quarter_dict:
value_list.append(last_five_quarter_dict[stock][1])
for i in range(4 - len(value_list)):
value_list.append(0)
tmp_dict['0Q'] = value_list[0]
tmp_dict['1Q'] = value_list[1]
tmp_dict['2Q'] = value_list[2]
tmp_dict['3Q'] = value_list[3]
tmp_dict['sum_value'] = value_list[0] + value_list[1] + value_list[2] + value_list[3]
tmp_list.append(tmp_dict)
df = pd.DataFrame(tmp_list)
return df
@classmethod
def fun_get_factor(cls, df, factor_name, industry, level, statsDate):
stock_list = BzUtil.fun_get_industry_stocks(industry, level, statsDate)
rs = BzUtil.fun_neutralize(stock_list, df, module=factor_name, industry_type=industry, level=level, statsDate=statsDate)
rs = BzUtil.fun_standardize(rs, 2)
return rs
@classmethod
def fun_diversity_by_industry(cls, stock_list, max_num, statsDate):
if stock_list is None:
return None
industry_list = BzUtil.fun_get_industry(cycle=None)
tmpList = []
for industry in industry_list:
i = 0
stocks = BzUtil.fun_get_industry_stocks(industry, 2, statsDate)
for stock in stock_list:
if stock in stocks: #by 行业选入 top max_num 的标的（如有）
i += 1
if i <= max_num:
tmpList.append(stock) #可能一个股票横跨多个行业，会导致多次入选，但不影响后面计算
final_stocks = []
for stock in stock_list:
if stock in tmpList:
final_stocks.append(stock)
return final_stocks
@classmethod
def fun_get_bad_stock_list(cls, statsDate):
#0、剔除商誉占比 > 30% 的股票
df = get_fundamentals(
query(valuation.code, balance.good_will, balance.equities_parent_company_owners),
date = statsDate
)
df = df.fillna(value = 0)
df['good_will_ratio'] = 1.0*df['good_will'] / df['equities_parent_company_owners']
list_good_will = list(df[df.good_will_ratio > 0.3].code)
bad_stocks = list_good_will
bad_stocks = list(set(bad_stocks))
return bad_stocks
@staticmethod
def get_inc_net_profile(statsDate):
# 取净利润增长率为正的
df = QuantLib.get_fundamentals_sum('income', income.net_profit, statsDate)
df = df.drop(['0Q', '1Q', '2Q', '3Q'], axis=1)
df.rename(columns={'sum_value':'ttm_1y'}, inplace=True)
df1 = QuantLib.get_fundamentals_sum('income', income.net_profit, (statsDate - dt.timedelta(365)))
df1 = df1.drop(['0Q', '1Q', '2Q', '3Q'], axis=1)
df1.rename(columns={'sum_value':'ttm_2y'}, inplace=True)
df = df.merge(df1, on='code')
df = df.fillna(value=0)
df['inc_net_profit'] = 1.0*(df['ttm_1y'] - df['ttm_2y'])
df = df[df.inc_net_profit > 0]
inc_net_profit_list = list(df.code)
return inc_net_profit_list
@staticmethod
def get_inc_operating_revenue_list(statsDate):
# 按行业取营业收入增长率前 1/3
df = QuantLib.get_fundamentals_sum('income', income.operating_revenue, statsDate)
df = df.drop(['0Q', '1Q', '2Q', '3Q'], axis=1)
df.rename(columns={'sum_value':'ttm_1y'}, inplace=True)
df1 = QuantLib.get_fundamentals_sum('income', income.operating_revenue, (statsDate - dt.timedelta(365)))
df1 = df1.drop(['0Q', '1Q', '2Q', '3Q'], axis=1)
df1.rename(columns={'sum_value':'ttm_2y'}, inplace=True)
df = df.merge(df1, on='code')
df = df.fillna(value=0)
df['inc_operating_revenue'] = 1.0*(df['ttm_1y'] - df['ttm_2y']) / abs(df['ttm_2y'])
df = df.fillna(value = 0)
industry_list = BzUtil.fun_get_industry(cycle=None)
inc_operating_revenue_list = []
for industry in industry_list:
stock_list = BzUtil.fun_get_industry_stocks(industry, 2, statsDate)
df_inc_operating_revenue = df[df.code.isin(stock_list)]
df_inc_operating_revenue = df_inc_operating_revenue.sort_values("inc_operating_revenue",ascending= False)
inc_operating_revenue_list = inc_operating_revenue_list + list(df_inc_operating_revenue[:int(len(df_inc_operating_revenue)*0.33)].code)
return inc_operating_revenue_list
@staticmethod
def get_low_liability_ratio(statsDate):
# 指标剔除资产负债率相对行业最高的1/3的股票
df = get_fundamentals(query(balance.code, balance.total_liability, balance.total_assets), date = statsDate)
df = df.fillna(value=0)
df['liability_ratio'] = 1.0*(df['total_liability'] / df['total_assets'])
industry_list = BzUtil.fun_get_industry(cycle=None)
liability_ratio_list = []
for industry in industry_list:
stock_list = BzUtil.fun_get_industry_stocks(industry, 2, statsDate)
df_liability_ratio = df[df.code.isin(stock_list)]
df_liability_ratio = df_liability_ratio.sort_values('liability_ratio', ascending=True)
liability_ratio_list = liability_ratio_list + list(df_liability_ratio[:int(len(df_liability_ratio)*0.66)].code)
return liability_ratio_list
@staticmethod
def get_high_profit_ratio(statsDate):
# 剔除净利润率相对行业最低的1/3的股票；
df = get_fundamentals(query(indicator.code, indicator.net_profit_to_total_revenue), date = statsDate)
df = df.fillna(value=0)
industry_list = BzUtil.fun_get_industry(cycle=None)
profit_ratio_list = []
for industry in industry_list:
stock_list = BzUtil.fun_get_industry_stocks(industry, 2, statsDate)
df_profit_ratio = df[df.code.isin(stock_list)]
df_profit_ratio = df_profit_ratio.sort_values('net_profit_to_total_revenue', ascending=False)
profit_ratio_list = profit_ratio_list + list(df_profit_ratio[:int(len(df_profit_ratio)*0.66)].code)
return profit_ratio_list
class ValueFactorLib():
def __init__(self):
pass
@staticmethod
def fun_get_stock_list(now_date, hold_number=10, statsDate=None):
relative_ps = ValueFactorLib.fun_get_relative_ps(statsDate)
low_ps = ValueFactorLib.fun_get_low_ps(statsDate)
good_stock_list = BzUtil.filter_intersection(relative_ps,low_ps)
# 取净利润增长率为正的
inc_net_profit_list = QuantLib.get_inc_net_profile(statsDate)
good_stock_list = BzUtil.filter_intersection(good_stock_list,inc_net_profit_list)
print((len(good_stock_list)))
# 按行业取营业收入增长率前 1/3
inc_operating_revenue_list = QuantLib.get_inc_operating_revenue_list(statsDate)
good_stock_list = list(set(good_stock_list) & set(inc_operating_revenue_list))
print((len(good_stock_list)))
# 指标剔除资产负债率相对行业最高的1/3的股票
liability_ratio_list = QuantLib.get_low_liability_ratio(statsDate)
good_stock_list = BzUtil.filter_intersection(good_stock_list,liability_ratio_list)
# 剔除净利润率相对行业最低的1/3的股票；
profit_ratio_list = QuantLib.get_high_profit_ratio(statsDate)
good_stock_list = BzUtil.filter_intersection(good_stock_list,profit_ratio_list)
stock_list = []
for stock in relative_ps:
if stock in good_stock_list:
stock_list.append(stock)
print((len(good_stock_list)))
stock_list = BzUtil.fun_delNewShare(now_date, stock_list, 30)
bad_stock_list = QuantLib.fun_get_bad_stock_list(statsDate)
stock_list = stock_list[:hold_number*10]
stock_list = BzUtil.filter_without(stock_list, bad_stock_list)
stock_list = BzUtil.remove_limit_up(stock_list)
stock_list = QuantLib.fun_diversity_by_industry(stock_list, int(hold_number*0.4), statsDate)
return stock_list[:hold_number]
@staticmethod
def fun_get_relative_ps(statsDate=None):
def __fun_get_ps(statsDate, deltamonth):
__df = get_fundamentals(query(valuation.code, valuation.ps_ratio), date = (statsDate - dt.timedelta(30*deltamonth)))
__df.rename(columns={'ps_ratio':deltamonth}, inplace=True)
return __df
for i in range(48):
df1 = __fun_get_ps(statsDate, i)
if i == 0:
df = df1
else:
df = df.merge(df1, on='code')
df.index = list(df['code'])
df = df.drop(['code'], axis=1)
df = df.fillna(value=0, axis=0)
# 1. 计算相对市收率，相对市收率等于个股市收率除以全市场的市收率，这样处理的目的是为了剔除市场估值变化的影响
for i in range(len(df.columns)):
s = df.iloc[:,i]
median = s.median()
df.iloc[:,i] = s / median
length, stock_list, stock_dict = len(df), list(df.index), {}
# 2. 计算相对市收率N个月的移动平均值的N个月的标准差，并据此计算布林带上下轨（N个月的移动平均值+/-N个月移动平均的标准差）。N = 24
for i in range(length):
s = df.iloc[i,:]
if s.min() < 0:
pass
else:
# tmp_list 是24个月的相对市收率均值
tmp_list = []
for j in range(24):
tmp_list.append(s[j:j+24].mean())
# mean_value 是最近 24个月的相对市收率均值
mean_value = tmp_list[0]
# std_value 是相对市收率24个月的移动平均值的24个月的标准差
std_value = np.std(tmp_list)
tmp_dict = {}
# (mean_value - std_value)，是布林线下轨（此处定义和一般布林线不一样，一般是 均线 - 2 倍标准差）
'''
研报原始的策略，选择 s[0] < mean_value - std_value 的标的，但因为 ps_ratio十分不稳定，跳跃很大，此区间里的测试结果非常不稳定
本策略退而求其次，选择均线-1倍标准差 和 均线 - 2 倍标准差之间的标的
大致反映策略的有效性
'''
if s[0] > (mean_value - 2.0*std_value) and s[0] < mean_value:
# 记录 相对市收率均值 / 当期相对市收率
stock_dict[stock_list[i]] = (1.0*mean_value/s[0])
stock_list = []
dict_score = stock_dict
dict_score = sorted(list(dict_score.items()), key=lambda d:d[1], reverse=True)
for idx in dict_score:
stock = idx[0]
stock_list.append(stock)
return stock_list
@staticmethod
def fun_get_low_ps(statsDate=None):
df = get_fundamentals(
query(valuation.code, valuation.ps_ratio),
date = statsDate
)
# 根据 sp 去极值、中性化、标准化后，跨行业选最佳的标的
industry_list = BzUtil.fun_get_industry(cycle=None)
df = df.fillna(value = 0)
sp_ratio = {}
df['SP'] = 1.0/df['ps_ratio']
df = df.drop(['ps_ratio'], axis=1)
for industry in industry_list:
tmpDict = QuantLib.fun_get_factor(df, 'SP', industry, 2, statsDate).to_dict()
for stock in tmpDict:
if stock in sp_ratio:
sp_ratio[stock] = max(sp_ratio[stock],tmpDict[stock])
else:
sp_ratio[stock] = tmpDict[stock]
dict_score = sorted(list(sp_ratio.items()), key=lambda d:d[1], reverse=True)
stock_list = []
for idx in dict_score:
stock = idx[0]
stock_list.append(stock)
return stock_list[:int(len(stock_list)*0.5)]
class Fisher():
def __init__(self):
self.factorlib = ValueFactorLib()
def fun_get_tdy_stock_list(self, current_dt):
now_date = current_dt
statsDate = now_date - datetime.timedelta(1)
stocks = ValueFactorLib.fun_get_stock_list(now_date=now_date,statsDate=statsDate)
return stocks
# 导入函数库
import statsmodels.api as sm
import random
class RSRSLib():
class RsPair(dict):
def __init__(self,ans=[],r2s=[],date='2005-01-01'):
self.ans = ans
self.r2s = r2s
self.date = date
def __getattr__(self, key):
try:
return self[key]
except KeyError:
raise AttributeError(r"'RsPair' object has no attribute '%s'" % key)
def __setattr__(self, key, value):
import datetime
if isinstance(value,datetime.date):
value = str(value)
self[key] = value
def __init__(self,buy_ratio=0.7,sell_ratio=-0.7):
self.rsrses = {}
self.buy = buy_ratio
self.sell = sell_ratio
def to_file(self,path='history_rsrs.json'):
write_file(path, str(self.rsrses))
def load_rsrs_from(self,path='history_rsrs.json'):
self.rsrses = {}
try:
data = read_file(path)
str_data = str(data,'utf-8')
str_data = str_data.replace("'", '"')
dict_data= json.loads(str_data)
rsrses = {}
for d in dict_data:
rs = self.RsPair(**dict_data[d])
rsrses[d] = rs
self.rsrses = rsrses
except Exception as e:
log.info('加载rsrs文件失败。'+str(e))
@staticmethod
def rsrs(prices):
highs = prices.high
lows = prices.low
X = sm.add_constant(lows)
model = sm.OLS(highs, X)
fit = model.fit()
# print('fit params:',fit.params)
beta = fit.params.low
#计算r2
r2=fit.rsquared
return beta,r2
def calc_rsrs_last(self,prices,N=18):
ans = []
r2s = []
for i in range(len(prices))[N:]:
parts = prices.iloc[i-N+1:i+1]
beta,r2 = self.rsrs(parts)
ans.append(beta)
#计算r2
r2s.append(r2)
return ans,r2s
def calc_zscore_rightdev(self,section,beta,r2):
# 计算均值序列
mu = np.mean(section)
# 计算标准化RSRS指标序列
sigma = np.std(section)
zscore = (section[-1]-mu)/sigma
#计算右偏RSRS标准分
zscore_rightdev= zscore*beta*r2
return zscore_rightdev
def calc_sec_rsrs_from(self,security,begin='2005-01-05',end=None, N=18):
log.info('计算'+str(begin)+'日至'+str(end)+'日的RSRS斜率指标')
if str(begin) >= str(end):
return None,None
prices = get_price(security, begin, end, '1d', ['high', 'low'])
size = len(prices.dropna())
if size > 0:
return self.calc_rsrs_last(prices.dropna(),N)
return None,None
def init_sec_rsrs(self,sec,end_date):
log.info('init sec rsrs[%s].'%(sec))
if sec in self.rsrses:
rs = self.RsPair(**self.rsrses[sec])
else:
rs = self.RsPair()
self.rsrses[sec] = rs
# 计算2005年1月5日至回测开始日期的RSRS斜率指标
ans,ans_rightdev = self.calc_sec_rsrs_from(sec, rs['date'], end_date)
if ans and ans_rightdev:
self.rsrses[sec] = self.RsPair(ans=ans,r2s=ans_rightdev,date=end_date)
def is_sec_buy_or_sell(self,sec,N=18,M=1100):
prices = attribute_history(sec, N, '1d', ['high', 'low'])
if len(prices.dropna()) == 0:
log.info('no data for sec'+sec)
return 'n'
beta, r2 = self.rsrs(prices.dropna())
self.rsrses[sec].ans.append(beta)
self.rsrses[sec].r2s.append(r2)
section = self.rsrses[sec].ans[-M:]
zscore_rightdev = self.calc_zscore_rightdev(section,beta,r2)
# 如果上一时间点的RSRS斜率大于买入阈值, 则全仓买入
if zscore_rightdev > self.buy:
# 记录这次买入
log.info("标准化RSRS斜率[%f]大于买入阈值, 买入 %s" % (zscore_rightdev,sec))
# 用所有 cash 买入股票
return 'b'
# 如果上一时间点的RSRS斜率小于卖出阈值, 则空仓卖出
if zscore_rightdev < self.sell:
# 记录这次卖出
log.info("标准化RSRS斜率[%f]小于卖出阈值, 卖出 %s" % (zscore_rightdev,sec))
# 卖出所有股票,使这只股票的最终持有量为0
return 's'
log.info("股票[%s]标准化RSRS斜率[%f]中性。" % (sec,zscore_rightdev))
return 'n' # 不处理
def is_hs300_buy(self):
return self.is_sec_buy_or_sell('000300.XSHG')
def judge_today_buy_or_sell(self,sec,now_date,N=18,M=1100):
self.init_sec_rsrs(sec,now_date)
return self.is_sec_buy_or_sell(sec,N,M)
def is_hs300_buy_tdy(self,now_date):
return self.judge_today_buy_or_sell('000300.XSHG',now_date)
class StopManager():
# 1 是否止损
# 2 止损记录
# 3 一段时间内不再购买
# 4 按先后排序
def __init__(self):
self.stop_ratio = 0.9 # 跌10%止损
self.stop_ndays = 20
self.blacks = {}
self.sorted_blacks = []
def check_stop(self,context):
self.context = context
for s in context.portfolio.positions:
p = context.portfolio.positions[s]
self.try_close(p)
def try_close(self, p):
# p:Position对象
if self.is_stop(p):
log.info('股票[%s]发生止损[%f,%f]。'%(p.security,p.price,p.avg_cost))
order_target(p.security, 0)
self.record(p.security)
def is_stop(self, position):
# position:Position对象
return position.price <= self.stop_ratio * position.avg_cost
def is_lost(self, position):
return position.price <= position.avg_cost
def record(self,sec):
# 记录sec,date
self.blacks[sec] = self.context.current_dt
if sec in self.sorted_blacks:
self.sorted_blacks.remove(sec)
self.sorted_blacks.append(sec)
def beyond_last_stop(self,stock,current_dt):
import datetime
stop_day = self.blacks[stock]
beyond_day = stop_day + datetime.timedelta(self.stop_ndays)
log.info('当前日期：'+str(current_dt)+' 逾期日：'+str(beyond_day))
return current_dt > beyond_day
def sort_by_stop_time(self,stocks):
sorted_stocks = []
tmp_stocks = stocks[::]
if len(tmp_stocks) == 0:
return sorted_stocks
for s in self.sorted_blacks:
if s in tmp_stocks:
sorted_stocks.append(s)
tmp_stocks.remove(s)
if len(tmp_stocks) == 0:
break
return sorted_stocks
def filter_and_sort(self,stocks,current_dt):
filted_stocks = []
need_sort = []
for s in stocks:
if s not in self.blacks:
filted_stocks.append(s)
if s in self.blacks:
log.info('股票[%s]发生过止损[%s]。'%(s,str(self.blacks[s])))
if self.beyond_last_stop(s,current_dt):
need_sort.append(s)
sorted_stocks = self.sort_by_stop_time(need_sort)
return filted_stocks + sorted_stocks
class RiskLib:
@staticmethod
def __get_daily_returns(stock_or_list, freq, lag):
hStocks = history(lag, freq, 'close', stock_or_list, df=True)
dailyReturns = hStocks.resample('D').last().pct_change().fillna(value=0, method=None, axis=0).values
return dailyReturns
@staticmethod
def __level_to_probability(confidencelevel):
# 正太分布标准差的倍数对应的分布概率
a = (1 - 0.95)
if confidencelevel == 1.96:
a = (1 - 0.95)
elif confidencelevel == 2.06:
a = (1 - 0.96)
elif confidencelevel == 2.18:
a = (1 - 0.97)
elif confidencelevel == 2.34:
a = (1 - 0.98)
elif confidencelevel == 2.58:
a = (1 - 0.99)
elif confidencelevel == 5:
a = (1 - 0.99999)
return a
@staticmethod
def calc_stock_ES(stock, a=0.05, freq='1d', lag=120):
ES = 0
fac = lag * a
dailyReturns = RiskLib.__get_daily_returns(stock, freq, lag)
dailyReturns_sort = sorted(dailyReturns)
count = 0
sum_value = 0
for i in range(len(dailyReturns_sort)):
if i < fac:
sum_value += dailyReturns_sort[i]
count += 1
if count > 0:
ES = -(sum_value / fac)
return ES[0]
@staticmethod