Residual Fund Flow Intensity Factor
factor.formula
Fund flow strength calculation formula (applicable to small and large orders):
Residual capital flow intensity calculation formula:
in:
- :
The capital flow intensity at time t. This formula is applicable to calculating the capital flow intensity of small orders and large orders. In the subsequent regression analysis, there will be two different $S_{small,t}$ for the capital flow intensity of small orders and the capital flow intensity of large orders.
- :
The buy transaction amount at time i is calculated separately for small orders and large orders, that is, there will be small order buy transaction amount ($Buy_{small,i}$) and large order buy transaction amount ($Buy_{large,i}$)
- :
The selling turnover at time i is calculated separately for small orders and large orders, that is, there will be small order selling turnover ($Sell_{small,i}$) and large order selling turnover ($Sell_{large,i}$)
- :
The time window for calculating the strength of capital flow. The default value is 1, and it can also be other positive integers, such as 10, 20, etc. Generally, it is a smaller value.
- :
The cumulative rate of return over the past 20 trading days at time t is calculated as follows: $Ret_{t,20} = \frac{P_t - P_{t-20}}{P_{t-20}}$, where $P_t$ is the price at time t
- :
The constant term of the linear regression model represents the expected capital flow intensity level when $Ret_{t,20}$ is 0
- :
The regression coefficient of the linear regression model represents the expected change in the intensity of capital flow ($S_t$) for each unit change in the cumulative return ($Ret_{t,20}$) over the past 20 days. The positive or negative value of $\beta$ represents the direction of the correlation between the intensity of capital flow and the return over the past 20 days.
- :
The residual term at time t represents the portion of the capital flow intensity that cannot be explained by the past 20-day yield ($Ret_{t,20}$). A positive residual indicates that the capital flow intensity is higher than the expected value based on the past 20-day yield, while a negative residual indicates that the capital flow intensity is lower than the expected value. The final residual capital flow intensity is the residual term $\varepsilon_t$
factor.explanation
The residual fund flow intensity factor is designed to strip away the impact of market price momentum on fund flow intensity. By linearly regressing the large and small order fund flow intensity with the cumulative returns of the past 20 days, the residual term represents the excess part of the fund flow intensity at the same return level. This factor can be used to identify stocks with independent fund flow signals. For example, when the residual fund flow intensity of a stock is positive, it indicates that the stock still attracts more than expected fund inflows under the current price momentum, which may indicate that the stock has the potential to rise in the future. On the contrary, a negative residual fund flow intensity suggests that fund outflows may be higher than expected based on price momentum, which may mean that the stock is at risk of falling in the future. This factor can be used alone or in combination with other factors to build a more robust quantitative stock selection model.