-
Notifications
You must be signed in to change notification settings - Fork 0
/
L3.1_FactorModel.py
186 lines (140 loc) · 7 KB
/
L3.1_FactorModel.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
#!/usr/bin/env python3
# -*- coding: utf-8 -*-
"""
Created on Tue Dec 21 21:25:09 2021
This code performs the following tasks:
1. load or download equity data from Yahoo and factor return data from Ken French Website
2. run regression models of the equity returns vs. the factor returns, from 1 factor to 4 factors
3. Export the model coefficient to an excel file
@author: fanyang
"""
#from pandas_datareader.famafrench import get_available_datasets
#dsa = get_available_datasets()
import pandas_datareader.data as web
import pandas as pd
#pd.core.common.is_list_like = pd.api.types.is_list_like
from pandas_datareader import data as pdr
import fix_yahoo_finance as yf
yf.pdr_override()
import datetime
import statsmodels.api as sm
import statsmodels.formula.api as smf
from statsmodels.sandbox.regression.predstd import wls_prediction_std
import matplotlib.pyplot as plt
import numpy as np
#function to get the price data from yahoo finance
def getDataBatch(tickers, startdate, enddate):
def getData(ticker):
return(pdr.get_data_yahoo(ticker, startdate, enddate))
datas = map(getData, tickers)
return(pd.concat(datas, keys=tickers, names=['tickers','Date']))
#------------------------------------------------------------------------
flag_downloadData = False # If True download data from web. Otherwise read in from the excel.
if flag_downloadData:
# define the time period
start_dt = datetime.datetime(2000, 1, 1)
end_dt = datetime.datetime(2018,12,31)
# get fama french data using pandas data reader
# fama french 3 factor model
ds = web.DataReader('F-F_Research_Data_Factors_daily','famafrench',
start = start_dt, end=end_dt)
df_FamaFrench = ds[0]
df_FamaFrench.columns = ['MKT', 'SMB', 'HML', 'RF']
df_FamaFrench['RF'] = 100*df_FamaFrench['RF']
# momentum factor
ds2 = web.DataReader('F-F_Momentum_Factor_daily','fammafrench',
start=start_dt,end=end_dt)
df2 = ds2[0]
df2.columns = ['MoM']
# combine the factors together
df_FamaFrench = pd.concat([df_FamaFrench,df2],axis = 1)
#---------------------------------------------------
# get stock price and find returns
# for one stock case,
stocks = pdr.get_data_yahoo('^GSPC', start=start_dt, end=end_dt)
# calculate returns
ret_stocks = stocks['Adj Close'].pct_change().dropna()
# for multiple stocks cases,
tickers = ['AAPL', 'MSFT', 'IBM', 'GOOG', '^GSPC', 'C', 'GE', 'PG', 'CELG', 'DIOD', 'FCNTX']
all_data = getDataBatch(tickers, start_dt, end_dt)
# Isolalte the 'Adj Close' values and transform the DataFrame
daily_close_px = all_data[['Adj Close']].reset_index().pivot('Date','Ticker','Adj Close')
# calculate the daily percentage change for 'daily_close_px'
daily_pct_change = daily_close_px.pct_change().dropna()
daily_pct_change.cloumns = ['AAPL', 'MSFT', 'IBM', 'GOOG', '^GSPC', 'C', 'GE', 'PG', 'CELG', 'DIOD', 'FCNTX']
# merge the return with famafrench
df_all = pd.concat([daily_pct_change*100,df_FamaFrench],axis = 1).dropna()
# subtract risk free rate from factors
# risk free rate needs to be divided by 250 because the data is daily
df_all_lessRF = df_all.copy()
# df_all['MKT'] = df_all['MKT']-1/250*df_all['RF']
for dfnames in df_all_lessRF.columns:
if dfnames not in ['MoM','SMB', 'HML', 'RF']:
df_all_lessRF[dfnames]=df_all_lessRF[dfnames]-1/250*df_all['RF']
# save data to excel
writer = pd.ExcelWriter('InputforRegression.xlsx',engine='xlsxwriter')
daily_pct_change.to_excel(writer,'StockReturn')
df_FamaFrench.to_excel(writer,'FactorReturn')
df_all.to_excel(writer,'AllReturn')
df_all_lessRF.to_excel(writer,'AllReturnLessRF')
writer.save()
else:
daily_pct_change = pd.read_excel('InputforRegression.xlsx',sheet_name = 'StockReturn')
df_FamaFrench = pd.read_excel('InputforRegression.xlsx',sheet_name='FactorReturn')
df_all = pd.read_excel('InputforRegression.xlsx',sheet_name='AllReturn')
df_all_lessRF = pd.read_excel('InputforRegression.xlsx',sheet_name='AllReturnLessRF')
#--------------------------------------------
# Linear Regression, CAPM model, using SP as market proxy
#----------------APPL-----------------------------
aapl_1f = smf.ols(formula='AAPL~MKT',data = df_all_lessRF).fit()
print(aapl_1f.summary())
aapl_1f_params = aapl_1f.params
aapl_1f_standardError = aapl_1f.bse
aapl_1f_PredValue = aapl_1f.predict()
aapl_1f_FittedValue = aapl_1f.fittedvalues
aapl_1f_Residual = aapl_1f.resid
aapl_1f_SumSqaureResidual = aapl_1f.ssr
#plot the data together with the linear regression fit
ret_MKT_sorted = np.sort(df_all_lessRF['MKT'])
#calculate the model 'predicted returns'for Apple given the market factor returns
ret_AAPL_sorted_calc = aapl_1f_params[0] + aapl_1f_params[1]*ret_MKT_sorted
plt.figure()
plt.plot(df_all_lessRF['MKT'],df_all_lessRF['AAPL'],'o',label='raw data')
plt.plot(ret_MKT_sorted, ret_AAPL_sorted_calc,'-r',label='CAPM model fit')
plt.xlabel('Market Returns', fontsize=12)
plt.ylabel('Apple Returns', fontsize=12)
plt.legend()
aapl_3f = smf.ols(formula = 'AAPL ~ MKT+SMB+HML',data = df_all_lessRF).fit()
print( aapl_3f.summary())
aapl_4f = smf.ols(formula = 'AAPL ~ MKT+SMB+HML+MoM',data = df_all_lessRF).fit()
print( aapl_4f.summary())
#--------------value stock-------------------------
pg_3f = smf.ols(formula='PG ~ MKT+SMB+HML', data=df_all_lessRF).fit()
print(pg_3f.summary())
pg_4f = smf.ols(formula='PG ~ MKT+SMB+HML+MoM', data=df_all_lessRF).fit()
print(pg_4f.summary())
#--------------Fidelity Contra Fund----------------
contra_3f = smf.ols(formula='Contra~MKT+SMB+HML', data=df_all_lessRF).fit()
print(contra_3f.summary())
contra_4f = smf.ols(formula = 'Contra~MKT+SMB+HML+MoM', data=df_all_lessRF).fit()
print(contra_4f.summary())
# Save data to a dataframe to be exported to excel files
Name_factors_save = ['Intercept','MKT','SMB','HML','MoM']
Name_stocks_save = ['AAPL','PG','Contra']
Parameters_save = np.vstack((aapl_4f.params.transpose(),pg_4f.params.transpose(),contra_4f.params.transpose()))
df_save = pd.DataFrame(Parameters_save,columns=Name_factors_save,index=Name_stocks_save)
# Write out results
writer=pd.ExcelWriter('Output_Coefficients.xlsx', engine='xlsxwriter')
df_save.to_excel(writer,sheet_name='FactorModel',
startrow=1, startcol=0, header=True, index=True)
# get the xlsxwriter object from the dataframe writer object
workbook = writer.book
# define desired formats for excel cells
cell_format_bold = workbook.add_format({'bold':True,'italic':True,'font_color':'red'})
cell_format1 = workbook.add_format({'font_color':'black','num_format':'0.00'})
worksheet1 = writer.sheets['FactorModel']
worksheet1.write('A1','Historical Beta Coefficients',cell_format_bold)
worksheet1.conditional_format('B3:F4',{'type':'cell',
'criteria':'>',
'value':-10000,
'format':cell_format1})