Beta Coefficient

Hi All,

I have been thinking about Beta Coefficient. Currently Beta’s is calculated in numerous ways as in CAPM. But I feel this mode of calculating beta may be flawed since the upside and downside volatilities differ. My opinion is that this should be calculated by regressing using the upside returns and down side return separately, i.e., my feeling is that the regression using upward returns and downward returns would be different. A stocks exposure to the market would be different in the upward movement and downward movement. Some stocks may react differently on downward movements than upward movements of the market, especially glamour, tech and growth stocks.

Any comments folks?

Best regards, Suminda Sirinath Salpitkorala Dharmasena