Sentiment Beta
factor.formula
Regression model of individual stock returns and market sentiment index:
Emotional sensitivity factor calculation formula:
In the formula:
- :
The daily return of stock i on day t is usually calculated as (closing price of the day - closing price of the previous day) / closing price of the previous day.
- :
The market sentiment index value on day t. The index can be composed of multiple market sentiment indicators, such as turnover rate, trading volume, price limit ratio, public opinion index, etc.
- :
The market sentiment index value on day t-1.
- :
The regression intercept term of stock i represents the expected return of the stock when the market sentiment index remains unchanged.
- :
The sentiment beta of stock i is obtained through time series regression, which measures the impact of changes in the market sentiment index on stock returns. A positive value indicates that stock returns tend to rise when market sentiment rises, while a negative value indicates the opposite.
- :
The residual term of the regression model represents the part of the return of stock i on day t that cannot be explained by the model.
factor.explanation
The sentiment sensitivity factor measures the impact of changes in market sentiment on the returns of individual stocks through a time-series regression model, and uses the absolute value of the Beta coefficient obtained by regression to negate it as the final factor value. The purpose of negating the absolute value is to make the factor value consistent with the risk preference, that is, the smaller the value, the greater the possibility that the individual stock will be negatively affected by market sentiment, and the higher the risk. Conversely, the larger the factor value, the less likely the individual stock is to be negatively affected by market sentiment, and the lower the risk. Therefore, this factor can be used to estimate risks and select stocks in quantitative investment. In practical applications, you can consider using a longer lookback window (such as more than 60 trading days) for time-series regression to improve model stability.