Normalized mean of large net buying strength during the opening period
factor.formula
in:
- :
Transform transaction data into buy and sell order data: Based on the order number of each transaction, the buy order and the sell order are matched to form the transaction data of the buy order and the sell order. This process is the basis for the subsequent large order identification.
- :
Definition of large order threshold: By statistically analyzing the transaction amount of buy and sell orders on historical trading days, the large order threshold is usually set by the method of "mean + k times standard deviation". For example, the logarithmically adjusted mean of the transaction amount of buy and sell orders in the past period of time plus 1 times the standard deviation can be used as the threshold for large order screening. The specific value of
kcan be adjusted according to market characteristics and backtesting results. This factor uses 1 times the standard deviation. - :
(NetTradeAmount_{i,n}): represents the net buy transaction amount of the i-th stock during the opening period (9:30-10:00) on the n-th trading day. The calculation method is:
large buy transaction amount - large sell transaction amount, where large orders are identified based on the threshold determined in step ②. - :
(\overline{NetTradeAmount}_{i,n}): represents the average of the net buying amount of the largest orders of the i-th stock during the opening period (9:30-10:00) on the nth trading day. There may be multiple trading minutes during the opening period. This value is the average of the
NetTradeAmountin these minutes. - :
(\sigma(NetTradeAmount_{i,n})): represents the standard deviation of the net buying amount of large orders during the opening period (9:30-10:00) of the ith stock on the nth trading day. It measures the volatility of the net buying of large orders during the opening period. The larger the standard deviation, the greater the volatility of the net buying of large orders during the opening period, and vice versa.
- :
(T): The length of the lookback period, which is the number of trading days used to calculate the factor value. For monthly stock selection strategies, it is usually set to 20 trading days; for weekly stock selection strategies, it is usually set to 5 trading days. A shorter lookback period captures short-term capital movements, while a longer lookback period takes into account longer-term capital inflow trends.
- :
This factor is calculated by the average value of
\frac{\overline{NetTradeAmount}_{i,n}}{\sigma(NetTradeAmount_{i,n})}on each trading day during the lookback period T. The mean of the numerator reflects the buying strength, the standard deviation of the denominator reflects the buying stability, and the final mean reflects the overall large-order net buying strength during the lookback period.
factor.explanation
This factor aims to capture the trading behavior of major funds during the opening hours by calculating the strength and stability of large net purchases during the opening hours over a period of time. Compared with only considering the net purchase amount of large orders, this factor measures both the strength and robustness of the purchase by standardizing the net purchase amount and calculating the average over a period of time. This factor can effectively identify stocks that are of interest to institutional investors, which usually have higher potential investment value. This factor can be combined with other technical and fundamental factors for multi-factor stock selection to improve the robustness of the stock selection strategy.