. use "600519.dta", clear
. describe
Contains data from d:\cnstock\beta\data/600519.dta
obs: 3,956
vars: 13 12 Apr 2018 09:17
size: 395,600
--------------------------------------------------------------
storage
variable name type variable label
--------------------------------------------------------------
stkcd long Stock Code
date double Trading Date
stknme str12 Stock Name
clsprc double Closing Price
hiprc double Highest Price
lowprc double Lowest Price
opnprc double Opening Price
rit double Daily Return
turnover double Turnover rate
volume long Trading Volume
transaction double Trading Amount in RMB
tcap double Total Market Capitalization
mcap double Circulation Market Capitalization
--------------------------------------------------------------
Sorted by: date
1.2 Stata 范例
global id="600519" //600519 贵州茅台
cntrade $id
global nn = stknme[1] //股票简称
*-股价时序图
use "$id.dta", clear
tsset date //设定为时间序列变量
tsline clsprc , xlabel(,angle(60)) subtitle("$nn $id")
graph export "$out\price_$id.png", replace
输出图片为：
贵州茅台-600519-股价走势-Stata连享会.png
2. 下载指数数据
2.1 基本下载方法
只需在 cntrade 命令后附加 index 选项即可下载如下指数：
000001 The Shanghai Composite Index (上证综合指数).
000300 CSI 300 Index (沪深 300).
399001 Shenzhen Component Index (深成指).
例如：
. cntrade 399001, index
d:\cnstock\beta\data
File tempcsvfile.csv (text file)
file 399001.dta has been saved
. use "399001.dta", clear
. describe
Contains data from 399001.dta
obs: 6,587
vars: 10 12 Apr 2018 09:26
size: 526,960
----------------------------------------------------
storage
variable name type variable label
----------------------------------------------------
indexcd long Index Code
date double Trading Date
indexnme str12 Index Name
clsprc double Closing Price
hiprc double Highest Price
lowprc double Lowest Price
opnprc double Opening Price
rmt double Daily Return
volume double Trading Volume
transaction double Trading Amount in RMB
----------------------------------------------------
Sorted by: date
2.2 批量下载多个指数并做进一步处理
foreach i in "399001" "000001" "000300" {
cntrade `i', index
rename rmt retmkt_`i' //更该指数收益名称，便于与个股合并
save index_`i'.dta, replace
erase `i'.dta
}
2.3 Stata 范例：沪深 300 时序图
* 调用其中一个指数数据：
use index_000300.dta, clear // 沪深300指数
tsset date
tsline clsprc , xlabel(,angle(60)) subtitle("沪深300指数")
沪深300指数-Stata连享会
3. 合并个股和指数数据
use "index_399001.dta", clear
merge 1:1 date using "$id.dta", nogen
gen year = year(date)
order date year index* stk* ret*
save "merge_$id.dta", replace
4. 估计 CAPM 模型
4.1 基本模型
use "merge_$id.dta", clear
*winsor2 rit, replace // winsor 处理
reg rit retmkt
结果如下：
. reg rit retmkt
Source | SS df MS Number of obs = 3,956
-------------+---------------------------------- F(1, 3954) = 1246.43
Model | .415684439 1 .415684439 Prob > F = 0.0000
Residual | 1.31865845 3,954 .0003335 R-squared = 0.2397
-------------+---------------------------------- Adj R-squared = 0.2395
Total | 1.73434289 3,955 .000438519 Root MSE = .01826
-------------------------------------------------------------------------------
rit | Coef. Std. Err. t P>|t| [95% Conf. Interval]
--------------+----------------------------------------------------------------
retmkt_399001 | 0.565 0.016 35.30 0.000 0.533 0.596
_cons | 0.001 0.000 4.42 0.000 0.001 0.002
-------------------------------------------------------------------------------
4.2 散点图和拟合情况
#delimit ;
twoway
(scatter rit retmkt if year<2008&abs(retmkt)<0.4,
xline(0,lp(dash) lc(red))
yline(0,lp(dash) lc(red)) msize(*0.2))
(lfit rit retmkt);
graph export "$out\scatter_$id.png", replace ;
#d cr
scatter_600519-Stata连享会.png
4.3 分时段估计
1998-2007
global yr = 2008
reg rit retmkt if year<$yr //&abs(retmkt)<0.4 // OLS
aaplot rit retmkt if e(sample), xline(0,lp(dash) lc(red)) ///
yline(0,lp(dash) lc(red)) msize(*0.6) ///
title("$nn ($id) 的 Beta 系数, 1998-2007") ///
xtitle("市场指数收益率(深成指)")
graph export "$out\aaplot_$id_b$yr.png", replace
*-下载 regfit 命令
. github install arlionn/regfit // 若已经有该命令，可以忽略本行命令
regfit, f(%4.2f) tvalue //呈现拟合方程
贵州茅台beta系数：1998-2007
. regfit, f(%4.2f) tvalue //呈现拟合方程
rit = 0.00 + 0.61*retmkt_399001
(3.90) (20.44)
N = 1466, R2 = 0.22, adj-R2 = 0.22
Note: t-value in parentheses
2008-2018
reg rit retmkt if year>=$yr //&abs(retmkt)<0.4 // OLS
aaplot rit retmkt if e(sample), xline(0,lp(dash) lc(red)) ///
yline(0,lp(dash) lc(red)) msize(*0.6) ///
title("$nn ($id) 的 Beta 系数, 2008-2017") ///
xtitle("市场指数收益率(深成指)")
graph export "$out\aaplot_$id_a$yr.png", replace
regfit, f(%4.2f) tvalue //呈现拟合方程
贵州茅台beta系数：2008-2018
. regfit, f(%4.2f) tvalue //呈现拟合方程
rit = 0.00 + 0.54*retmkt_399001
(2.44) (28.93)
N = 2490, R2 = 0.25, adj-R2 = 0.25
Note: t-value in parentheses
5. 分析个股 Beta 系数变化特征
5.1 逐年回归
. bysort year: reg rit retmkt, noheader
-------------------------------------------------------------------------------
-> year = 1998
no observations
-------------------------------------------------------------------------------
-> year = 1999
no observations
-------------------------------------------------------------------------------
-> year = 2000
no observations
-------------------------------------------------------------------------------
-> year = 2001
-------------------------------------------------------------------------------
rit | Coef. Std. Err. t P>|t| [95% Conf. Interval]
--------------+----------------------------------------------------------------
retmkt_399001 | 0.598 0.117 5.09 0.000 0.364 0.831
_cons | 0.004 0.002 1.65 0.102 -0.001 0.008
-------------------------------------------------------------------------------
(略)
-------------------------------------------------------------------------------
-> year = 2017
-------------------------------------------------------------------------------
rit | Coef. Std. Err. t P>|t| [95% Conf. Interval]
--------------+----------------------------------------------------------------
retmkt_399001 | 0.624 0.118 5.28 0.000 0.392 0.857
_cons | 0.003 0.001 2.90 0.004 0.001 0.005
-------------------------------------------------------------------------------
-------------------------------------------------------------------------------
-> year = 2018
-------------------------------------------------------------------------------
rit | Coef. Std. Err. t P>|t| [95% Conf. Interval]
--------------+----------------------------------------------------------------
retmkt_399001 | 0.731 0.158 4.64 0.000 0.416 1.047
_cons | 0.001 0.002 0.30 0.764 -0.004 0.005
-------------------------------------------------------------------------------
5.2 逐年回归: 更为简洁的处理方法
statsby _b[retmkt], by(year) saving("beta_data.dta", replace): ///
reg rit retmkt
变化趋势
*-变化趋势
preserve
use "beta_data.dta", clear
keep if _stat!=.
qui sum year
local byr = r(min)
local eyr = r(max)
list
#d ;
twoway connect _stat year,
yline(1, lpattern(dash) lcolor(red*0.3))
xlabel(`byr'(2)`eyr')
ylabel(,angle(0) format(%2.1f))
subtitle("$nn ($id) 的Beta系数")
ytitle("beta") xtitle("");
#d cr
graph export "$out\beta $nn $id.png", replace
restore
beta 贵州茅台 600519
附：完整命令
* =============================
*
* Stata 连享会: 证券分析系列
*
* 2. Beta 系数
*
* =============================
*-注意：执行后续命令之前，请先执行如下命令
global path "d:\cnstock\beta" //只需修改自路径即可，其他文件夹会自动生成
cd $path
cap mkdir data
cap mkdir refs
cap mkdir out
cap mkdir adofiles
global D "$path\data" //范例数据
global R "$path\refs" //参考文献
global out "$path\out" //图形和表格结果
adopath + "$path\adofiles" //添加外部命令所在路径
cd $D
set scheme s2color
/*
*------------------
*- 0 安装外部命令
*------------------
*-设定安装外部路径的位置
net set ado "$path\adofiles"
net set other "$D" //外部命令附带的 example datasets 等存放于此
*-安装外部命令
ssc install cntrade, replace //下载个股和市场指数实时数据, SJ 14(2):381--388
ssc install openall, replace //批量合并
ssc install aaplot, replace
ssc install regfit, replace
*/
*---------------------
*- 1 下载个股交易资料
*---------------------
*-请输入股票代码
*-股票代码查询
view browse "http://quote.stockstar.com/stock/stock_index.htm"
*-部分股票代码
/*
global id="000002" //000002 万科
global id="600028" //600028 中石化
global id="000527" //000527 美的电器,已经退市
global id="000651" //000651 格力电器
global id="000429" //000429 粤高速A
global id="000541" //000541 佛山照明
global id="000596" //000596 古井贡酒
global id="000799" //000799 酒鬼酒
global id="600048" //600048 保利地产 2006年07月31日
global id="600115" //600115 东方航空
global id="600085" //600085 同仁堂
global id="600249" //600249 两面针
global id="300398" //300398 飞凯材料 2014年10月09日
global id="603799" //603799 华友钴业,2015年01月29日
global id="300431" //300431 暴风影音, 2015年03月24日
*/
global id="600519" //600519 贵州茅台
cntrade $id
global nn = stknme[1] //股票简称
use "$id.dta", clear
tsset date //设定为时间序列变量
*-股价时序图
tsline clsprc , xlabel(,angle(60)) subtitle("$nn $id")
graph export "$out\price_$id.png", replace
*------------------
*- 2 下载指数数据
*------------------
* -Index Codes and Index Names:
*
* 000001 The Shanghai Composite Index (上证综合指数).
* 000300 CSI 300 Index (沪深 300).
* 399001 Shenzhen Component Index (深成指).
foreach i in "399001" "000001" "000300" {
cntrade `i', index
rename rmt retmkt_`i'
save index_`i'.dta, replace
erase `i'.dta
}
/* 调用其中一个指数数据：
use index_000300.dta, clear // 沪深300指数
tsset date
tsline clsprc , xlabel(,angle(60)) subtitle("沪深300指数")
*/
*-----------------------
*- 3 合并个股和指数数据
*-----------------------
use "index_399001.dta", clear
merge 1:1 date using "$id.dta", nogen
gen year = year(date)
order date year index* stk* ret*
save "merge_$id.dta", replace
*---------------------------------
*- 4 估计CAPM模型并分析 Beta 系数
*---------------------------------
*-估计CAPM模型
use "merge_$id.dta", clear
*winsor2 reti, replace // winsor 处理
reg rit retmkt
#delimit ;
twoway
(scatter rit retmkt if year<2008&abs(retmkt)<0.4,
xline(0,lp(dash) lc(red))
yline(0,lp(dash) lc(red)) msize(*0.2))
(lfit rit retmkt);
#d cr
// 图示拟合情况
global yr = 2008
reg rit retmkt if year<$yr //&abs(retmkt)<0.4 // OLS
aaplot rit retmkt if e(sample), xline(0,lp(dash) lc(red)) ///
yline(0,lp(dash) lc(red)) msize(*0.6) ///
title("$nn ($id) 的 Beta 系数, 1998-2007") ///
xtitle("市场指数收益率(深成指)")
graph export "$out\aaplot_$id_b$yr.png", replace
regfit, f(%4.2f) tvalue //呈现拟合方程
reg rit retmkt if year>=$yr //&abs(retmkt)<0.4 // OLS
aaplot rit retmkt if e(sample), xline(0,lp(dash) lc(red)) ///
yline(0,lp(dash) lc(red)) msize(*0.6) ///
title("$nn ($id) 的 Beta 系数, 2008-2017") ///
xtitle("市场指数收益率(深成指)")
graph export "$out\aaplot_$id_a$yr.png", replace
regfit, f(%4.2f) tvalue //呈现拟合方程
*-逐年回归
bysort year: reg rit retmkt, noheader
*-逐年统计
statsby _b[retmkt], by(year) saving("beta_data.dta", replace): ///
reg rit retmkt
*-变化趋势
preserve
use "beta_data.dta", clear
keep if _stat!=.
qui sum year
local byr = r(min)
local eyr = r(max)
list
#d ;
twoway connect _stat year,
yline(1, lpattern(dash) lcolor(red*0.3))
xlabel(`byr'(2)`eyr')
ylabel(,angle(0) format(%2.1f))
subtitle("$nn ($id) 的Beta系数")
ytitle("beta") xtitle("");
#d cr
graph export "beta $nn $id.wmf", replace
restore
*--------------------------over------------------------
