Factors Directory

Quantitative Trading Factors

Analyst Coverage Residuals

Emotional FactorsFundamental factors

factor.formula

Analyst coverage residual calculation formula:

in:

  • :

    The analyst coverage of the i-th stock at the end of the m-th month can be selected as simple analyst coverage or total analyst coverage. This value indicates how many analysts covered stock i at the end of the m-th month, reflecting the market's attention to the stock. Adding 1 to the logarithm here is to avoid the situation where the logarithm cannot be taken when COV is 0. At the same time, it can also smooth the data distribution to a certain extent and reduce the impact of outliers.

  • :

    Take the natural logarithm of the total market value of the i-th stock at the end of the m-th month. The total market value reflects the size of the company. Generally, companies with larger market values ​​receive more analyst attention. Therefore, using the logarithm of market value as an explanatory variable can control the impact of company size on analyst coverage. Using the logarithmic form can also reduce the impact of extreme market values, making its data distribution more consistent with the assumptions of linear regression.

  • :

    is the natural logarithm of the average daily turnover rate of the ith stock in the past three months up to the end of the mth month. The turnover rate reflects the liquidity of the stock, and stocks with high liquidity are usually more popular in the market. The purpose of taking the logarithm is to reduce the impact of extreme turnover rates and make their distribution more stable.

  • :

    is the return of the i-th stock over the past three months up to the end of the m-th month. The momentum effect of stocks indicates that stocks that have performed well in the past may continue to perform well in the future, so analysts tend to pay attention to such stocks. The returns over the past three months are used here to capture the momentum effect as a control variable for analyst coverage.

  • :

    The intercept term of the regression model represents the expected value of the logarithm of analyst coverage when all explanatory variables are 0. In practical terms, it represents the baseline level of analyst coverage when market capitalization, turnover rate, and momentum effects are not considered.

  • :

    The regression coefficient of the logarithm of market capitalization ($SIZE_{i,m}$) represents the expected change in the logarithm of analyst coverage for each unit change in the logarithm of market capitalization. This coefficient reflects the direction and degree of the impact of stock size on analyst coverage, and is expected to be positive, that is, the larger the market capitalization of the company, the higher its analyst coverage.

  • :

    The regression coefficient of the logarithm of the daily average turnover rate in the past three months ($LNTO_{i,m}$) represents the expected change in the logarithm of analyst coverage for every unit change in the logarithm of the turnover rate. This coefficient reflects the direction and degree of the impact of stock liquidity on analyst coverage, and is expected to be positive, that is, the more liquid the stock, the higher its analyst coverage.

  • :

    The regression coefficient of the past three-month return ($MOM_{i,m}$) represents the expected change in the logarithm of analyst coverage for every unit change in the return. This coefficient reflects the direction and degree of the impact of the stock momentum effect (past performance) on analyst coverage, and is expected to be positive, that is, the stock with good past performance has higher analyst coverage.

  • :

    The residual term of the regression model represents the analyst coverage that cannot be explained by market value, turnover rate and momentum effect, i.e., analyst coverage residual. This residual term is considered to be abnormal analyst coverage, which may reflect the analyst's selective bias, information advantage and other behaviors, and is also the focus of this factor.

factor.explanation

This factor decomposes analyst coverage into two parts through regression analysis: one part is the expected analyst coverage that can be explained by the basic characteristics of the stock (such as market capitalization, turnover rate and momentum effect); the other part is the regression residual, which represents the abnormal analyst coverage beyond expectations, that is, the analyst coverage residual. Studies have shown that the excess returns of stocks are significantly correlated with analyst coverage residuals, reflecting factors such as selectivity bias and information advantage in analyst behavior, which cannot be fully explained by market capitalization, turnover rate and momentum effect. Therefore, this factor captures information asymmetry and analyst behavioral bias, and has certain alpha mining potential.

Related Factors