Volume Percentile Reversal Strength Factor
factor.formula
1. Calculate the 13/16th percentile of the daily transaction amount distribution. This percentile represents the higher level of transaction amount distribution on that day, which can reflect the activity of large transactions.
2. Select the 10 trading days with the highest 13/16 percentile values, and calculate the arithmetic sum of the gains and losses of these 10 trading days, recorded as $M_{high}$. $M_{high}$ represents the cumulative return during the period of active large-scale transactions.
3. Select the 10 trading days with the lowest 13/16 percentile values, and calculate the arithmetic sum of the gains and losses of these 10 trading days, recorded as $M_{low}$. $M_{low}$ represents the cumulative return during the period of inactive large-scale transactions.
4. Calculate the turnover percentile reversal strength factor M: $M$ represents the difference in returns between active and inactive periods of large transactions. The larger the difference, the stronger the reversal signal.
in:
- :
The stock price change on the trading day with the highest 13/16th percentile value, where i=1,2,...,10.
- :
The stock price change on the jth trading day with the lowest 13/16 percentile value, where j=1,2,...,10.
- :
The arithmetic sum of the gains and losses in the 10 trading days with the highest 13/16 percentile values represents the cumulative returns during the period of active large-scale transactions.
- :
The arithmetic sum of the gains and losses in the 10 trading days with the lowest 13/16 percentile values represents the cumulative gains during the period of inactive large transactions.
- :
The turnover percentile reversal strength factor, whose value is equal to the difference between $M_{high}$ and $M_{low}$, reflects the difference in returns between periods of high turnover activity and periods of low turnover activity. The larger the difference, the stronger the reversal signal.
factor.explanation
This factor is based on the market microstructure theory and believes that changes in the turnover quantile contain rich market information. The early version of the ideal reversal factor used the average daily single transaction amount as a measurement standard, but ignored the skewness of the turnover distribution. This factor can more effectively capture the activities of large transactions by introducing the 13/16 quantile value of the intraday transaction amount distribution, and then extract stronger reversal signals. The logic of this factor is that when large transactions (high quantile values) occur frequently, it may indicate that market sentiment is extreme and the possibility of reversal is higher. On the contrary, when large transactions decrease, it may mean that market sentiment is stable or the direction is stabilizing. The core idea of this factor is to capture the micro-level game of market participants on prices and predict reversal opportunities in the short term through changes in the turnover distribution. This factor is suitable for highly liquid stocks and requires high-frequency trading data support.