<跟着小邢老师学python> 第八课生成夏普比率评价交易

有的时候, 不同的交易策略最终的收益是一样或者接近的, 但是其中的风险并不一致

夏普比率就是一个很好的的评价方式

数据

这里有两份交易数据 (都是我手动造出来的)

data1 = [['2016-07-04', 5000000.00], ['2016-07-05', 4835769.38], ['2016-07-06', 5036838.71], ['2016-07-07', 4881597.25], ['2016-07-08', 5099866.54], ['2016-07-11', 5158305.86], ['2016-07-12', 5379073.81], ['2016-07-13', 5174389.13], ['2016-07-14', 5235795.42], ['2016-07-15', 5050606.65], ['2016-07-18', 5153493.81], ['2016-07-19', 5410731.34], ['2016-07-20', 5233841.72], ['2016-07-21', 5245693.42], ['2016-07-22', 5089468.61], ['2016-07-25', 5368514.36], ['2016-07-26', 5640139.80], ['2016-07-27', 5285918.31], ['2016-07-28', 5669054.75], ['2016-07-29', 5954109.95], ['2016-08-01', 5363140.57], ['2016-08-02', 5463225.81], ['2016-08-03', 5535254.28], ['2016-08-04', 5580878.38], ['2016-08-05', 5583599.43], ['2016-08-08', 5532192.08], ['2016-08-09', 5651183.49], ['2016-08-10', 5668028.27], ['2016-08-11', 5595957.57], ['2016-08-12', 5602812.38], ['2016-08-15', 5668724.48], ['2016-08-16', 5617281.02], ['2016-08-17', 5715316.60], ['2016-08-18', 5770128.40], ['2016-08-19', 5732294.20], ['2016-08-22', 5751937.92], ['2016-08-23', 5874266.66], ['2016-08-24', 5853859.26], ['2016-08-25', 5880573.21], ['2016-08-26', 5890153.76], ['2016-08-29', 5894311.85], ['2016-08-30', 5852469.53], ['2016-08-31', 5834809.34], ['2016-09-01', 5937106.72], ['2016-09-02', 5848258.50], ['2016-09-05', 5835814.67], ['2016-09-06', 5919762.95], ['2016-09-07', 5945231.91], ['2016-09-08', 5996092.42]]
data2 = [['2016-07-04', 5000000.00], ['2016-07-05', 5035769.38], ['2016-07-06', 5036838.71], ['2016-07-07', 5081597.25], ['2016-07-08', 5099866.54], ['2016-07-11', 5158305.86], ['2016-07-12', 5179073.81], ['2016-07-13', 5174389.13], ['2016-07-14', 5135795.42], ['2016-07-15', 5150606.65], ['2016-07-18', 5153493.81], ['2016-07-19', 5210731.34], ['2016-07-20', 5233841.72], ['2016-07-21', 5245693.42], ['2016-07-22', 5289468.61], ['2016-07-25', 5368514.36], ['2016-07-26', 5440139.80], ['2016-07-27', 5485918.31], ['2016-07-28', 5469054.75], ['2016-07-29', 5454109.95], ['2016-08-01', 5463140.57], ['2016-08-02', 5463225.81], ['2016-08-03', 5535254.28], ['2016-08-04', 5580878.38], ['2016-08-05', 5583599.43], ['2016-08-08', 5532192.08], ['2016-08-09', 5651183.49], ['2016-08-10', 5668028.27], ['2016-08-11', 5595957.57], ['2016-08-12', 5602812.38], ['2016-08-15', 5668724.48], ['2016-08-16', 5617281.02], ['2016-08-17', 5715316.60], ['2016-08-18', 5770128.40], ['2016-08-19', 5732294.20], ['2016-08-22', 5751937.92], ['2016-08-23', 5874266.66], ['2016-08-24', 5853859.26], ['2016-08-25', 5880573.21], ['2016-08-26', 5890153.76], ['2016-08-29', 5894311.85], ['2016-08-30', 5852469.53], ['2016-08-31', 5834809.34], ['2016-09-01', 5937106.72], ['2016-09-02', 5848258.50], ['2016-09-05', 5835814.67], ['2016-09-06', 5919762.95], ['2016-09-07', 5945231.91], ['2016-09-08', 5996092.42]]

比较时间一致, 初始资产一致, 最终收益一致

夏普比率的公式

夏普比率 = 每天收益率的均值 / 每天收益率的标准差

这里省略了无风险利率

根据这个公式就能实现如下函数来计算

def calculate_sharp(data):
    df = pd.DataFrame(data, columns=['date', 'liquidate'])
    df['rate_of_return'] = df['liquidate'].rolling(2).apply(lambda x: x[1]/x[0] - 1, raw=True)

    sharp = df['rate_of_return'].mean() / df['rate_of_return'].std()
    return sharp

rolling是窗口函数,这里用来比较相邻的两天收益情况
mean是均值
std是标准差

结果

data1与data2的夏普比率分别是 0.137771和0.451409

可以看出第2组数据更好一些, 因为收益稳定,波动小一些

完整代码

# 计算夏普比率
import pandas as pd

def calculate_sharp(data):
    df = pd.DataFrame(data, columns=['date', 'liquidate'])
    df['rate_of_return'] = df['liquidate'].rolling(2).apply(lambda x: x[1]/x[0] - 1, raw=True)

    sharp = df['rate_of_return'].mean() / df['rate_of_return'].std()
    return sharp

data1 = [['2016-07-04', 5000000.00], ['2016-07-05', 4835769.38], ['2016-07-06', 5036838.71], ['2016-07-07', 4881597.25], ['2016-07-08', 5099866.54], ['2016-07-11', 5158305.86], ['2016-07-12', 5379073.81], ['2016-07-13', 5174389.13], ['2016-07-14', 5235795.42], ['2016-07-15', 5050606.65], ['2016-07-18', 5153493.81], ['2016-07-19', 5410731.34], ['2016-07-20', 5233841.72], ['2016-07-21', 5245693.42], ['2016-07-22', 5089468.61], ['2016-07-25', 5368514.36], ['2016-07-26', 5640139.80], ['2016-07-27', 5285918.31], ['2016-07-28', 5669054.75], ['2016-07-29', 5954109.95], ['2016-08-01', 5363140.57], ['2016-08-02', 5463225.81], ['2016-08-03', 5535254.28], ['2016-08-04', 5580878.38], ['2016-08-05', 5583599.43], ['2016-08-08', 5532192.08], ['2016-08-09', 5651183.49], ['2016-08-10', 5668028.27], ['2016-08-11', 5595957.57], ['2016-08-12', 5602812.38], ['2016-08-15', 5668724.48], ['2016-08-16', 5617281.02], ['2016-08-17', 5715316.60], ['2016-08-18', 5770128.40], ['2016-08-19', 5732294.20], ['2016-08-22', 5751937.92], ['2016-08-23', 5874266.66], ['2016-08-24', 5853859.26], ['2016-08-25', 5880573.21], ['2016-08-26', 5890153.76], ['2016-08-29', 5894311.85], ['2016-08-30', 5852469.53], ['2016-08-31', 5834809.34], ['2016-09-01', 5937106.72], ['2016-09-02', 5848258.50], ['2016-09-05', 5835814.67], ['2016-09-06', 5919762.95], ['2016-09-07', 5945231.91], ['2016-09-08', 5996092.42]]
data2 = [['2016-07-04', 5000000.00], ['2016-07-05', 5035769.38], ['2016-07-06', 5036838.71], ['2016-07-07', 5081597.25], ['2016-07-08', 5099866.54], ['2016-07-11', 5158305.86], ['2016-07-12', 5179073.81], ['2016-07-13', 5174389.13], ['2016-07-14', 5135795.42], ['2016-07-15', 5150606.65], ['2016-07-18', 5153493.81], ['2016-07-19', 5210731.34], ['2016-07-20', 5233841.72], ['2016-07-21', 5245693.42], ['2016-07-22', 5289468.61], ['2016-07-25', 5368514.36], ['2016-07-26', 5440139.80], ['2016-07-27', 5485918.31], ['2016-07-28', 5469054.75], ['2016-07-29', 5454109.95], ['2016-08-01', 5463140.57], ['2016-08-02', 5463225.81], ['2016-08-03', 5535254.28], ['2016-08-04', 5580878.38], ['2016-08-05', 5583599.43], ['2016-08-08', 5532192.08], ['2016-08-09', 5651183.49], ['2016-08-10', 5668028.27], ['2016-08-11', 5595957.57], ['2016-08-12', 5602812.38], ['2016-08-15', 5668724.48], ['2016-08-16', 5617281.02], ['2016-08-17', 5715316.60], ['2016-08-18', 5770128.40], ['2016-08-19', 5732294.20], ['2016-08-22', 5751937.92], ['2016-08-23', 5874266.66], ['2016-08-24', 5853859.26], ['2016-08-25', 5880573.21], ['2016-08-26', 5890153.76], ['2016-08-29', 5894311.85], ['2016-08-30', 5852469.53], ['2016-08-31', 5834809.34], ['2016-09-01', 5937106.72], ['2016-09-02', 5848258.50], ['2016-09-05', 5835814.67], ['2016-09-06', 5919762.95], ['2016-09-07', 5945231.91], ['2016-09-08', 5996092.42]]

sharp1 = calculate_sharp(data1)
sharp2 = calculate_sharp(data2)

print(sharp1, sharp2)

最后

我近期一边整理免费课程，也会推出更多的免费视频，方便大家结合查看学习

如果你对我的课程感兴趣，欢迎与我联系，提供一对一教学，也可以帮助实现特定程序

了解清楚目标后，第一次可以先听课后付款

如果有概念没有理解清楚，随时可以询问，不再收费

回到首页查看更多课程

我的联系方式

微信：ilovebcc （需要一对一教学请加好友时注明）
YY: 2341272648
QQ: 383124540

<跟着小邢老师学python> 第八课 生成夏普比率评价交易

数据

夏普比率的公式

结果

完整代码

最后

我的联系方式

<跟着小邢老师学python> 第八课生成夏普比率评价交易