阿尔法和贝塔计算

import pandas as pd
import numpy as np
from scipy import stats
import akshare as ak
def get_return(code,startdate,endate):
    df=ak.stock_zh_a_daily(code, start_date=startdate, end_date=endate, adjust="qfq")
    p1=np.array(df.close[1:])
    p0=np.array(df.close[:-1])
    logret = np.log(p1/p0)
    rate=pd.DataFrame()
    rate[code]=logret
    rate.index=df['date'][1:]
    return rate

收益率函数；从akshare下载数据；

导入系数：

startdate="20201103"
endate="20211103"
code="sz000905"
get_return(code,startdate,endate)

结果：

 

 一、阿尔法和贝塔计算

贝塔与中证500的收益率做对比，中证500股票代码：sz000905

定义阿尔法和贝塔函数：

def alpha_beta(code,startdate,endate):
    mkt_ret=get_return("sz000905",startdate,endate)
    stock_ret=get_return(code,startdate,endate)
    df=pd.merge(mkt_ret,stock_ret,left_index=True,right_index=True)
    x=df.iloc[:,0]
    y=df.iloc[:,1]
    beta,alpha,r_value,p_value,std_err=stats.linregress(x,y)
    return (alpha,beta)

导入系数：

startdate="20201103"
endate="20211103"
code="sh600519"
alpha_beta(code,startdate,endate)

结果：(0.0002538222146195239, -0.14144412140986778)

二、计算多个股票的阿尔法和贝塔

def stocks_alpha_beta(stocks,startdate,endate):
 
    df=pd.DataFrame()
    alpha=[]
    beta=[]
 
    for code in stocks.values():
        a,b=alpha_beta(code,startdate,endate)
        alpha.append(float('%.4f'%a))
        beta.append(float('%.2f'%b))
    df['alpha']=alpha
    df['beta']=beta
    df.index=stocks.keys()
    return df

导入系数：

startdate="20201103"
endate="20211103"
stocks={"中国平安":"sh601318","格力电器":"sh600036"}
stocks_alpha_beta(stocks,startdate,endate)

结果：