Pyfinanceを用いた証券リターン分析のためのPythonデータビジュアライゼーション 例 詳細

import pandas as pd  
import numpy as np
from pyfinance import TSeries
import tushare as ts
def get_data(code,start='2011-01-01',end=''):
    df=ts.get_k_data(code,start,end)
    df.index=pd.to_datetime(df.date)
    ret=df.close/df.close.shift(1)-1
    # return TSeries sequence
    return TSeries(ret.dropna())
# Get the Ping An of China data
tss=get_data('601318')
#tss.head()

#annualized return
anl_ret=tss.anlzd_ret()
#Cumulative return
cum_ret=tss.cuml_ret()
# Calculate the periodic rate of return
q_ret=tss.rollup('Q')
a_ret=tss.rollup('A')
print(f'Annualized return: {anl_ret*100:.2f}%')
print(f'Cumulative return: {cum_ret*100:.2f}%')
#print(f'Quarterly return: {q_ret.tail().round(4)}')
#print(f'Calendar year return: {a_ret.round(4)}')

from pyecharts import Bar
attr=q_ret.index.strftime('%Y%m')
v1=(q_ret*100).round(2).values
bar=Bar('Ping An of China quarterly return %') bar.add('',attr,v1,)
bar

from pyecharts import Bar
attr=a_ret.index.strftime('%Y')
v1=(a_ret*100).round(2).values
bar=Bar('Ping An of China calendar year return %')
bar.add('',attr,v1,is_label_show=True,
       is_splitline_show=False)
bar

#Based on the CSI 300 index
#To ensure the length of both is the same, the index of Ping An of China is used as the benchmark
benchmark=get_data('hs300')
benchmark=benchmark.loc[tss.index]
alpha,beta,rsq=tss.alpha(benchmark),tss.beta(benchmark),tss.rsq(benchmark)
tstat_a,tstat_b=tss.tstat_alpha(benchmark),ttss.tstat_beta(benchmark)
print(f'alpha:{alpha:.4f},tstat: {tstat_a:.2f}')
print(f'beta :{beta:.4f}, t-statistic: {tstat_b:.2f}')
print(f'regression coefficient of determination R2: {tss.rsq(benchmark):.3f}')

alpha :0.0004, t-statistic: 1.55
beta :1.0634, t-statistic: 60.09
Regression coefficient of determination R2: 0.606

#annualized standard deviation
a_std=tss.anlzd_stdev()
#downstream standard deviation
s_std=tss.semi_stdev()
#max retracement
md=tss.max_drawdown()
print(f'Annualized standard deviation: {a_std*100:.2f}%')
print(f'downward bias standard deviation: {s_std*100:.2f}%')
print(f'Maximum retracement deviation: {md*100:.2f}%')

Annualized standard deviation: 31.37%
Lower bias standard deviation: 0.43%
Maximum retracement difference: -45.76%

bat=tss.batting_avg(benchmark)
uc=tss.up_capture(benchmark)
dc=tss.down_capture(benchmark)
tc=uc/dc
pct_neg=tss.pct_negative()
pct_pos=tss.pct_positive()
print(f'Percentage of time above benchmark return: {bat*100:.2f}%')
print(f'Ratio of upward periods to benchmark returns: {uc*100:.2f}%')
print(f'Downside to benchmark return ratio: {dc*100:.2f}%')
print(f'Upside to Downside Ratio: {tc*100:.2f}%')
print(f'Individual stock downside (negative earnings) time ratio: {pct_neg*100:.2f}%')
print(f'Time share of individual stocks on the upside (positive returns): {pct_pos*100:.2f}%')

Percentage of time with higher returns than benchmark: 47.83%
Ratio of upward period to benchmark return: 111.70%
Ratio of downside period to benchmark return: 105.32%
Ratio of upside period to downside period: 106.06%
Individual downside (negative return) time ratio: 48.94%
Individual stock upside (positive return) time ratio: 50.00%

ir=tss.info_ratio(benchmark)
tr=tss.treynor_ratio(benchmark)
print(f'info ratio: {ir:.3f}')
print(f'Treynor index: {tr:.3f}')
Information ratio: 0.433
Traynor index: 0.096

sr=tss.sharpe_ratio()
sor=tss.sortino_ratio(freq=250)
cr=tss.calmar_ratio()
print(f'Sharpe ratio: {sr:.2f}')
print(f'Sortino ratio: {sor:.2f}')
print(f'Karma ratio: {cr:.2f}')

Sharpe ratio: 0.33
Sortino ratio: 28.35
Karma ratio: 0.27

def performance(code,start='2011-01-01',end=''):
    tss=get_data(code,start,end)
    benchmark=get_data('hs300',start,end).loc[tss.index]
    dd={}
    #Returns
    #annualized rate of return
    dd['annualized_yield']=tss.anlzd_ret()
    #Cumulative yield
    dd['cumulative_ret']=tss.cuml_ret()
    #alpha and beta
    dd['alpha']=tss.alpha(benchmark)
    dd['beta']=tss.beta(benchmark)
    #risk metrics
    #annualized standard deviation
    dd['annualized standard deviation']=tss.anlzd_stdev()
    #Downside standard deviation
    dd['downside standard deviation']=tss.semi_stdev()
    #Maximum retracement
    dd['max_drawdown']=tss.max_drawdown()
    #Information ratio and Traynor index
    dd['info_ratio']=tss.info_ratio(benchmark)
    dd['Treynor index']=tss.treynor_ratio(benchmark)
    #Risk-adjusted return
    dd['Sharpe ratio']=tss.sharpe_ratio()
    dd['sortino_ratio']=tss.sortino_ratio(freq=250)
    dd['calmar ratio']=tss.calmar_ratio()
    df=pd.DataFrame(dd.values(),index=dd.keys()).round(4)
    return df

#Get data for multiple stocks
df=pd.DataFrame(index=performance('601318').index)
stocks={'Ping An of China':'601318','Guizhou Maotai':'600519',\
        'Haitian Flavor':'603288','Gree Electric':'000651',\
        'Vanke A':'00002','BYD':'002594',\
        'Yunnan Baiyao':'000538','Shuanghui Development':'000895',\
        'Haier Zhijia':'600690','Tsingtao Beer':'600600'}
for name,code in stocks.items():
    try:
        df[name]=performance(code).values
    except:
        continue

d