p********0
发帖数: 186 | 1 Hi, I read AHXZ paper that calculate idiosyncratic risk for each stock using
Fama - French,
r_{i} = alpha_{I} + beta_{i}*MKT + a_{i} * SMB + b_{i} * HML +epsilon_{i},
we do a daily regression based past one month 20 data points to get the
variance(epsilon_{i}), here i is the index of the stock so each stock has
its own coefficient _beta_{i}, a_{i}, b_{i}.
Do we need to have constraint for small cap copanies the a_{i} need to be >0
and value companies b_{i} need to postive??? Otherwise the idiosyncratic
risk may not be accurate in case a small company behave more like a large
cap company. | b*******r
发帖数: 28 | 2 no
using
>0
【在 p********0 的大作中提到】
: Hi, I read AHXZ paper that calculate idiosyncratic risk for each stock using
: Fama - French,
: r_{i} = alpha_{I} + beta_{i}*MKT + a_{i} * SMB + b_{i} * HML +epsilon_{i},
: we do a daily regression based past one month 20 data points to get the
: variance(epsilon_{i}), here i is the index of the stock so each stock has
: its own coefficient _beta_{i}, a_{i}, b_{i}.
: Do we need to have constraint for small cap copanies the a_{i} need to be >0
: and value companies b_{i} need to postive??? Otherwise the idiosyncratic
: risk may not be accurate in case a small company behave more like a large
: cap company.
|
|
