学习记录655@python双均线突破量化回测实战-以比亚迪为例

教练我想学编程

已于 2022-10-05 14:41:31 修改

阅读量1k

点赞数

文章标签： python 量化回测

于 2022-10-05 14:19:05 首次发布

本文链接：https://blog.csdn.net/weixin_44663675/article/details/127172013

版权

以下代码对比亚迪2020年1月1日至2022年10月5日进行均线策略回测
策略逻辑是:如果5日均线大于10日均线,那么就持有一手仓位,否则不持有仓位,就这么简单明了,然后换算成净值,看净值图情况,查看回测的效果。
根据净值图，最高净值可达到最初净值的20倍。
ps：现实情况肯定没这么简单，比如核心要解决的问题就是选股先选到比亚迪，其次是坚定不宜的持有这么久，才能得到这么高的收益。

import tushare as ts
ts.set_token('你的token')
# 以比亚迪前复权日线数据为例进行回测
df = ts.pro_bar(ts_code='002594.SZ', adj='qfq', start_date='20200101', end_date='20221004')
df.sort_values(by="trade_date",inplace=True)
df

	ts_code	trade_date	open	high	low	close	pre_close	change	pct_chg	vol	amount
667	002594.SZ	20200102	47.6292	48.4082	47.4495	48.1086	47.6093	0.4993	1.0487	159345.70	765516.493
666	002594.SZ	20200103	48.1386	48.9276	47.6192	47.9788	48.1086	-0.1298	-0.2698	129936.07	628361.961
665	002594.SZ	20200106	47.3296	49.1273	47.1299	48.2185	47.9788	0.2397	0.4996	169871.38	822172.472
664	002594.SZ	20200107	48.2484	48.4382	47.7091	47.9888	48.2185	-0.2297	-0.4764	93400.58	449013.435
663	002594.SZ	20200108	47.4894	48.2884	47.0899	47.2198	47.9888	-0.7690	-1.6025	110974.45	529249.404
...	...	...	...	...	...	...	...	...	...	...	...
4	002594.SZ	20220926	266.4000	276.0000	263.5800	272.9500	266.4000	6.5500	2.4587	143606.40	3911414.127
3	002594.SZ	20220927	272.9600	274.8300	270.0600	273.9600	272.9500	1.0100	0.3700	83095.93	2262822.665
2	002594.SZ	20220928	273.0000	273.0000	265.5000	265.6200	273.9600	-8.3400	-3.0442	85620.18	2293531.516
1	002594.SZ	20220929	270.0000	271.8000	264.0000	264.5700	265.6200	-1.0500	-0.3953	88356.68	2355802.796
0	002594.SZ	20220930	263.0000	263.5500	249.0000	252.0100	264.5700	-12.5600	-4.7473	214035.88	5421385.039

668 rows × 11 columns

import talib as ta

# 两条均线的参数
L1=5
L2=10

# 使用talib计算移动均线
df['ma1']=ta.SMA(df.close.values,timeperiod=L1)
df['ma2']=ta.SMA(df.close.values,timeperiod=L2)
df

	ts_code	trade_date	open	high	low	close	pre_close	change	pct_chg	vol	amount	ma1	ma2
667	002594.SZ	20200102	47.6292	48.4082	47.4495	48.1086	47.6093	0.4993	1.0487	159345.70	765516.493	NaN	NaN
666	002594.SZ	20200103	48.1386	48.9276	47.6192	47.9788	48.1086	-0.1298	-0.2698	129936.07	628361.961	NaN	NaN
665	002594.SZ	20200106	47.3296	49.1273	47.1299	48.2185	47.9788	0.2397	0.4996	169871.38	822172.472	NaN	NaN
664	002594.SZ	20200107	48.2484	48.4382	47.7091	47.9888	48.2185	-0.2297	-0.4764	93400.58	449013.435	NaN	NaN
663	002594.SZ	20200108	47.4894	48.2884	47.0899	47.2198	47.9888	-0.7690	-1.6025	110974.45	529249.404	47.9029	NaN
...	...	...	...	...	...	...	...	...	...	...	...	...	...
4	002594.SZ	20220926	266.4000	276.0000	263.5800	272.9500	266.4000	6.5500	2.4587	143606.40	3911414.127	270.4220	272.107
3	002594.SZ	20220927	272.9600	274.8300	270.0600	273.9600	272.9500	1.0100	0.3700	83095.93	2262822.665	270.3720	271.023
2	002594.SZ	20220928	273.0000	273.0000	265.5000	265.6200	273.9600	-8.3400	-3.0442	85620.18	2293531.516	269.4960	269.752
1	002594.SZ	20220929	270.0000	271.8000	264.0000	264.5700	265.6200	-1.0500	-0.3953	88356.68	2355802.796	268.7000	269.259
0	002594.SZ	20220930	263.0000	263.5500	249.0000	252.0100	264.5700	-12.5600	-4.7473	214035.88	5421385.039	265.8220	267.862

668 rows × 13 columns

# 计算趋势，0代表没有趋势不做操作，空仓，1代表多头趋势，第二日买入
df['trend']=0
con_long=df['ma1']>df['ma2']
con_short=df['ma1']<=df['ma2']
df.loc[con_long,'trend']=1
df.loc[con_short,'trend']=0
df

	ts_code	trade_date	open	high	low	close	pre_close	change	pct_chg	vol	amount	ma1	ma2	trend
667	002594.SZ	20200102	47.6292	48.4082	47.4495	48.1086	47.6093	0.4993	1.0487	159345.70	765516.493	NaN	NaN	0
666	002594.SZ	20200103	48.1386	48.9276	47.6192	47.9788	48.1086	-0.1298	-0.2698	129936.07	628361.961	NaN	NaN	0
665	002594.SZ	20200106	47.3296	49.1273	47.1299	48.2185	47.9788	0.2397	0.4996	169871.38	822172.472	NaN	NaN	0
664	002594.SZ	20200107	48.2484	48.4382	47.7091	47.9888	48.2185	-0.2297	-0.4764	93400.58	449013.435	NaN	NaN	0
663	002594.SZ	20200108	47.4894	48.2884	47.0899	47.2198	47.9888	-0.7690	-1.6025	110974.45	529249.404	47.9029	NaN	0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
4	002594.SZ	20220926	266.4000	276.0000	263.5800	272.9500	266.4000	6.5500	2.4587	143606.40	3911414.127	270.4220	272.107	0
3	002594.SZ	20220927	272.9600	274.8300	270.0600	273.9600	272.9500	1.0100	0.3700	83095.93	2262822.665	270.3720	271.023	0
2	002594.SZ	20220928	273.0000	273.0000	265.5000	265.6200	273.9600	-8.3400	-3.0442	85620.18	2293531.516	269.4960	269.752	0
1	002594.SZ	20220929	270.0000	271.8000	264.0000	264.5700	265.6200	-1.0500	-0.3953	88356.68	2355802.796	268.7000	269.259	0
0	002594.SZ	20220930	263.0000	263.5500	249.0000	252.0100	264.5700	-12.5600	-4.7473	214035.88	5421385.039	265.8220	267.862	0

668 rows × 14 columns

# 假设出现信号后，第二天开盘进行交易，每次开仓1手(即100股)
# 注意pos代表目标仓位,如果trend>1,也就是当5日均线大于10日均线,就持有一手仓位,否则补充持有仓位
# new_pos为当日新增仓位,old_pos为老仓位
# 新旧仓位这里不好理解,可按照表格自己计算梳理下
df['pos']=100*df['trend'].shift(1)
df['pos_shift']=df['pos'].shift(1)
df['new_pos']=df['pos']-df['pos'].shift(1)
df['old_pos']=df['pos']-df['new_pos']
df

	ts_code	trade_date	open	high	low	close	pre_close	change	pct_chg	vol	amount	ma1	ma2	trend	pos	pos_shift	new_pos	old_pos
667	002594.SZ	20200102	47.6292	48.4082	47.4495	48.1086	47.6093	0.4993	1.0487	159345.70	765516.493	NaN	NaN	0	NaN	NaN	NaN	NaN
666	002594.SZ	20200103	48.1386	48.9276	47.6192	47.9788	48.1086	-0.1298	-0.2698	129936.07	628361.961	NaN	NaN	0	0.0	NaN	NaN	NaN
665	002594.SZ	20200106	47.3296	49.1273	47.1299	48.2185	47.9788	0.2397	0.4996	169871.38	822172.472	NaN	NaN	0	0.0	0.0	0.0	0.0
664	002594.SZ	20200107	48.2484	48.4382	47.7091	47.9888	48.2185	-0.2297	-0.4764	93400.58	449013.435	NaN	NaN	0	0.0	0.0	0.0	0.0
663	002594.SZ	20200108	47.4894	48.2884	47.0899	47.2198	47.9888	-0.7690	-1.6025	110974.45	529249.404	47.9029	NaN	0	0.0	0.0	0.0	0.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
4	002594.SZ	20220926	266.4000	276.0000	263.5800	272.9500	266.4000	6.5500	2.4587	143606.40	3911414.127	270.4220	272.107	0	0.0	0.0	0.0	0.0
3	002594.SZ	20220927	272.9600	274.8300	270.0600	273.9600	272.9500	1.0100	0.3700	83095.93	2262822.665	270.3720	271.023	0	0.0	0.0	0.0	0.0
2	002594.SZ	20220928	273.0000	273.0000	265.5000	265.6200	273.9600	-8.3400	-3.0442	85620.18	2293531.516	269.4960	269.752	0	0.0	0.0	0.0	0.0
1	002594.SZ	20220929	270.0000	271.8000	264.0000	264.5700	265.6200	-1.0500	-0.3953	88356.68	2355802.796	268.7000	269.259	0	0.0	0.0	0.0	0.0
0	002594.SZ	20220930	263.0000	263.5500	249.0000	252.0100	264.5700	-12.5600	-4.7473	214035.88	5421385.039	265.8220	267.862	0	0.0	0.0	0.0	0.0

668 rows × 18 columns

# 把开盘价作为交易价格
df['entry_p']=df['open']
df

	ts_code	trade_date	open	high	low	close	pre_close	change	pct_chg	vol	amount	ma1	ma2	trend	pos	pos_shift	new_pos	old_pos	entry_p
667	002594.SZ	20200102	47.6292	48.4082	47.4495	48.1086	47.6093	0.4993	1.0487	159345.70	765516.493	NaN	NaN	0	NaN	NaN	NaN	NaN	47.6292
666	002594.SZ	20200103	48.1386	48.9276	47.6192	47.9788	48.1086	-0.1298	-0.2698	129936.07	628361.961	NaN	NaN	0	0.0	NaN	NaN	NaN	48.1386
665	002594.SZ	20200106	47.3296	49.1273	47.1299	48.2185	47.9788	0.2397	0.4996	169871.38	822172.472	NaN	NaN	0	0.0	0.0	0.0	0.0	47.3296
664	002594.SZ	20200107	48.2484	48.4382	47.7091	47.9888	48.2185	-0.2297	-0.4764	93400.58	449013.435	NaN	NaN	0	0.0	0.0	0.0	0.0	48.2484
663	002594.SZ	20200108	47.4894	48.2884	47.0899	47.2198	47.9888	-0.7690	-1.6025	110974.45	529249.404	47.9029	NaN	0	0.0	0.0	0.0	0.0	47.4894
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
4	002594.SZ	20220926	266.4000	276.0000	263.5800	272.9500	266.4000	6.5500	2.4587	143606.40	3911414.127	270.4220	272.107	0	0.0	0.0	0.0	0.0	266.4000
3	002594.SZ	20220927	272.9600	274.8300	270.0600	273.9600	272.9500	1.0100	0.3700	83095.93	2262822.665	270.3720	271.023	0	0.0	0.0	0.0	0.0	272.9600
2	002594.SZ	20220928	273.0000	273.0000	265.5000	265.6200	273.9600	-8.3400	-3.0442	85620.18	2293531.516	269.4960	269.752	0	0.0	0.0	0.0	0.0	273.0000
1	002594.SZ	20220929	270.0000	271.8000	264.0000	264.5700	265.6200	-1.0500	-0.3953	88356.68	2355802.796	268.7000	269.259	0	0.0	0.0	0.0	0.0	270.0000
0	002594.SZ	20220930	263.0000	263.5500	249.0000	252.0100	264.5700	-12.5600	-4.7473	214035.88	5421385.039	265.8220	267.862	0	0.0	0.0	0.0	0.0	263.0000

668 rows × 19 columns

# 计算旧仓位的盈利和新仓位的盈利
# 新仓位当日盈亏为当日收盘价与开盘价之差乘以新仓位
# 旧仓位当日盈亏为当日收盘价与昨日收盘价之差乘以旧仓位
df['p&l_new']=(df['close']-df['entry_p'])*df['new_pos']
df['p&l_old']=(df['close']-df['close'].shift(1))*df['old_pos']
# 交易费用,new_pos为100或者-100表示有交易,交易费用统一为5元
df['fee']=0
df.loc[(df['new_pos']==100) | (df['new_pos']==-100),'fee']=5
# 每日盈亏由三部分组成
df['p&l']=df['p&l_new']+df['p&l_old']-df['fee']
df

	ts_code	trade_date	open	high	low	close	pre_close	change	pct_chg	vol	...	trend	pos	pos_shift	new_pos	old_pos	entry_p	p&l_new	p&l_old	fee	p&l
667	002594.SZ	20200102	47.6292	48.4082	47.4495	48.1086	47.6093	0.4993	1.0487	159345.70	...	0	NaN	NaN	NaN	NaN	47.6292	NaN	NaN	0	NaN
666	002594.SZ	20200103	48.1386	48.9276	47.6192	47.9788	48.1086	-0.1298	-0.2698	129936.07	...	0	0.0	NaN	NaN	NaN	48.1386	NaN	NaN	0	NaN
665	002594.SZ	20200106	47.3296	49.1273	47.1299	48.2185	47.9788	0.2397	0.4996	169871.38	...	0	0.0	0.0	0.0	0.0	47.3296	0.0	0.0	0	0.0
664	002594.SZ	20200107	48.2484	48.4382	47.7091	47.9888	48.2185	-0.2297	-0.4764	93400.58	...	0	0.0	0.0	0.0	0.0	48.2484	-0.0	-0.0	0	0.0
663	002594.SZ	20200108	47.4894	48.2884	47.0899	47.2198	47.9888	-0.7690	-1.6025	110974.45	...	0	0.0	0.0	0.0	0.0	47.4894	-0.0	-0.0	0	0.0
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
4	002594.SZ	20220926	266.4000	276.0000	263.5800	272.9500	266.4000	6.5500	2.4587	143606.40	...	0	0.0	0.0	0.0	0.0	266.4000	0.0	0.0	0	0.0
3	002594.SZ	20220927	272.9600	274.8300	270.0600	273.9600	272.9500	1.0100	0.3700	83095.93	...	0	0.0	0.0	0.0	0.0	272.9600	0.0	0.0	0	0.0
2	002594.SZ	20220928	273.0000	273.0000	265.5000	265.6200	273.9600	-8.3400	-3.0442	85620.18	...	0	0.0	0.0	0.0	0.0	273.0000	-0.0	-0.0	0	0.0
1	002594.SZ	20220929	270.0000	271.8000	264.0000	264.5700	265.6200	-1.0500	-0.3953	88356.68	...	0	0.0	0.0	0.0	0.0	270.0000	-0.0	-0.0	0	0.0
0	002594.SZ	20220930	263.0000	263.5500	249.0000	252.0100	264.5700	-12.5600	-4.7473	214035.88	...	0	0.0	0.0	0.0	0.0	263.0000	-0.0	-0.0	0	0.0

668 rows × 23 columns

# 计算累计盈亏
df['p&l_cum']=df['p&l'].cumsum()

# 计算净值曲线，假设初始资金是1000
ini_cap=1000
df['capital']=df['p&l_cum']+ini_cap
df['net_value']=df['capital']/ini_cap
df['net_value']=df.net_value.apply(lambda x:x if x>=0 else 0)
df

	ts_code	trade_date	open	high	low	close	pre_close	change	pct_chg	vol	...	new_pos	old_pos	entry_p	p&l_new	p&l_old	fee	p&l	p&l_cum	capital	net_value
667	002594.SZ	20200102	47.6292	48.4082	47.4495	48.1086	47.6093	0.4993	1.0487	159345.70	...	NaN	NaN	47.6292	NaN	NaN	0	NaN	NaN	NaN	0.00000
666	002594.SZ	20200103	48.1386	48.9276	47.6192	47.9788	48.1086	-0.1298	-0.2698	129936.07	...	NaN	NaN	48.1386	NaN	NaN	0	NaN	NaN	NaN	0.00000
665	002594.SZ	20200106	47.3296	49.1273	47.1299	48.2185	47.9788	0.2397	0.4996	169871.38	...	0.0	0.0	47.3296	0.0	0.0	0	0.0	0.00	1000.00	1.00000
664	002594.SZ	20200107	48.2484	48.4382	47.7091	47.9888	48.2185	-0.2297	-0.4764	93400.58	...	0.0	0.0	48.2484	-0.0	-0.0	0	0.0	0.00	1000.00	1.00000
663	002594.SZ	20200108	47.4894	48.2884	47.0899	47.2198	47.9888	-0.7690	-1.6025	110974.45	...	0.0	0.0	47.4894	-0.0	-0.0	0	0.0	0.00	1000.00	1.00000
...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...	...
4	002594.SZ	20220926	266.4000	276.0000	263.5800	272.9500	266.4000	6.5500	2.4587	143606.40	...	0.0	0.0	266.4000	0.0	0.0	0	0.0	14668.52	15668.52	15.66852
3	002594.SZ	20220927	272.9600	274.8300	270.0600	273.9600	272.9500	1.0100	0.3700	83095.93	...	0.0	0.0	272.9600	0.0	0.0	0	0.0	14668.52	15668.52	15.66852
2	002594.SZ	20220928	273.0000	273.0000	265.5000	265.6200	273.9600	-8.3400	-3.0442	85620.18	...	0.0	0.0	273.0000	-0.0	-0.0	0	0.0	14668.52	15668.52	15.66852
1	002594.SZ	20220929	270.0000	271.8000	264.0000	264.5700	265.6200	-1.0500	-0.3953	88356.68	...	0.0	0.0	270.0000	-0.0	-0.0	0	0.0	14668.52	15668.52	15.66852
0	002594.SZ	20220930	263.0000	263.5500	249.0000	252.0100	264.5700	-12.5600	-4.7473	214035.88	...	0.0	0.0	263.0000	-0.0	-0.0	0	0.0	14668.52	15668.52	15.66852

668 rows × 26 columns

# 绘制净值曲线
df=df.set_index('trade_date')
df.plot(figsize=(12,6),y=['net_value'])

<AxesSubplot:xlabel='trade_date'>

[外链图片转存失败,源站可能有防盗链机制,建议将图片保存下来直接上传(img-RxLeXcoi-1664949829950)(output_8_1.png)]
如上图所示，最高净值涨幅近20倍。

教练我想学编程

关注

0
点赞
踩
8

收藏

觉得还不错? 一键收藏
0
评论
学习记录655@python双均线突破量化回测实战-以比亚迪为例

策略逻辑是:如果5日均线大于10日均线,那么就持有一手仓位,否则不持有仓位,就这么简单明了,然后换算成净值,看净值图情况,查看回测的效果。ps：现实情况肯定没这么简单，比如核心要解决的问题就是选股先选到比亚迪，其次是坚定不宜的持有这么久，才能得到这么高的收益。以下代码对比亚迪2020年1月1日至2022年10月5日进行均线策略回测。根据净值图，最高净值可达到最初净值的20倍。如上图所示，最高净值涨幅近20倍。
复制链接

扫一扫