Influence factor determination method and device, storage medium, electronic device and computer program product

Abstract

The invention discloses an influence factor determination method and device, a storage medium, an electronic device and a computer program product, and relates to the field of computers.The influence factor determination method comprises the steps that target historical data of a target transaction system is acquired, the target historical data comprises a historical evaluation price, a historical transaction price and a historical transaction evaluation value; according to the target historical data, N initial influence factors are determined, the initial influence factors are factors influencing the transaction price of the target transaction system, and N is an integer greater than or equal to 2; the N initial influence factors are screened, M target influence factors are obtained, the transaction price, determined according to the M target influence factors, of the target transaction system is larger than or equal to the transaction price, determined according to any M initial influence factors in the N initial influence factors, of the target transaction system, and the transaction price of the target transaction system is larger than or equal to the transaction price, determined according to any M initial influence factors in the N initial influence factors, of the target transaction system. M is a positive integer smaller than or equal to N.

Description

Technical Field

[0001] The present application relates to the field of computers, and more specifically, to a method and device for determining an influencing factor, a storage medium, an electronic device, and a computer program product. Background Art

[0002] Intelligent decision-making systems in the computer field are developing rapidly, particularly in supply chain management, e-commerce, and intelligent trading systems. Using data-driven analytics to improve decision-making efficiency and accuracy has become a research hotspot. However, intelligent trading systems currently face the challenge of mining and analyzing influencing factors. Traditional analytical methods for dealing with such trading systems primarily rely on single-factor models, such as item popularity and transaction time periods. These methods suffer from various issues, including single-factor design and a lack of specificity: Traditional analysis typically relies on a single factor, making it difficult to fully capture the characteristics and market dynamics of goods or services in the trading system; non-standardized factor selection processes and low predictive accuracy: Factor selection relies heavily on subjective judgment and lacks effective quantitative verification, resulting in inaccurate and unstable price predictions for goods or services in the trading system; and low factor calculation efficiency: Due to the diverse range of goods involved in trading systems and the complex data dimensions, traditional data processing and calculation methods struggle to efficiently integrate and process large amounts of data, limiting the depth and breadth of analysis. Consequently, it is difficult to accurately and efficiently identify the factors that most influence trading systems.

[0003] Currently, no effective solution has been proposed to the problem of low accuracy and efficiency in determining factors that affect trading systems in related technologies.

[0004] Therefore, it is necessary to improve the related technology to overcome the above-mentioned defects in the related technology. Summary of the Invention

[0005] The embodiments of the present application provide a method and device for determining an influencing factor, a storage medium, an electronic device, and a computer program product to at least solve the problem of low accuracy and efficiency in determining factors that affect a trading system.

[0006] According to one aspect of an embodiment of the present application, a method for determining an influencing factor is provided, comprising: obtaining target historical data of a target trading system, wherein the target historical data comprises: historical evaluation prices, historical transaction prices, and historical transaction evaluation values; determining N initial influencing factors based on the target historical data, wherein the initial influencing factors are factors that affect the transaction price of the target trading system, and N is an integer greater than or equal to 2; screening the N initial influencing factors to obtain M target influencing factors, wherein the transaction price of the target trading system determined based on the M target influencing factors is greater than or equal to the transaction price of the target trading system determined based on any M initial influencing factors among the N initial influencing factors, and M is a positive integer less than or equal to N.

[0007] In an exemplary embodiment, the N initial influencing factors are screened to obtain M target influencing factors, including: determining N information coefficients and N information coefficient ratios based on the N initial influencing factors, wherein the i-th information coefficient among the N information coefficients is the information coefficient between the i-th initial influencing factor among the N initial influencing factors and the historical transaction price, and the i-th information coefficient ratio among the N information coefficient ratios is the information coefficient ratio between the i-th initial influencing factor and the historical transaction price, the information coefficient is used to measure the correlation between the influencing factor and the historical transaction price, and the information coefficient ratio is used to measure the stability of the correlation, and i is a positive integer less than or equal to N; determining P candidate influencing factors from the N initial influencing factors based on the N information coefficients and the N information coefficient ratios, wherein the information coefficient of each candidate influencing factor among the P candidate influencing factors is greater than a first preset threshold value, and the information coefficient ratio is greater than a second preset threshold value, and P is a positive integer less than or equal to N; and determining the M target influencing factors from the P candidate influencing factors.

[0008] In an exemplary embodiment, the N initial influencing factors are screened to obtain M target influencing factors, including: dividing the N initial influencing factors into X groups of influencing factors according to a first preset classification condition, wherein each group of influencing factors in the X groups of influencing factors includes at least one initial influencing factor, and X is a positive integer less than or equal to N; dividing each group of influencing factors in the X groups of influencing factors into a plurality of small groups according to a second preset classification condition, to obtain Y small groups, wherein each group of the Y small groups includes at least one initial influencing factor, and Y is an integer greater than or equal to X; calculating a transaction price corresponding to each of the Y small groups based on the initial influencing factors in each of the Y small groups; using the initial influencing factors in one or more target groups as candidate influencing factors to obtain Z candidate influencing factors, wherein the transaction price corresponding to the target group is greater than a third preset threshold, and Z is a positive integer less than or equal to N; and determining the M target influencing factors from the Z candidate influencing factors.

[0009] In an exemplary embodiment, when W is equal to P or Z, the M target influencing factors are determined from the W candidate influencing factors, including: using a forward step screening method and / or a backward stepwise elimination method and / or a stability selection method to determine the M target influencing factors from the W candidate influencing factors.

[0010] In an exemplary embodiment, after obtaining M target impact factors, the method further includes: dividing a preset time period to obtain E sub-time periods, where E is a positive integer; determining the transaction price of the target trading system in each sub-time period of the E sub-time periods to obtain E transaction prices, wherein the transaction price of the e-th sub-time period of the E sub-time periods is determined based on the M target impact factors and the weights corresponding to the M target impact factors, each of the M target impact factors has a different corresponding weight in different sub-time periods, and e is a positive integer less than or equal to E; calculating the target indicator value corresponding to the target trading system in the preset time period based on the E transaction prices, wherein the target indicator value includes: rate of return, Sharpe ratio, and maximum drawdown value; when the rate of return is less than a fourth preset threshold, and / or the Sharpe ratio is less than a fifth preset threshold, and / or the maximum drawdown value is greater than a sixth preset threshold, the N initial impact factors are screened again to obtain multiple target impact factors.

[0011] In an exemplary embodiment, after obtaining M target influencing factors, the method further includes: adjusting the transaction conditions of the target trading system, wherein the transaction conditions include: the type of trading goods and the transaction duration; determining the transaction price of the target trading system under different transaction conditions based on the M target influencing factors; and when the transaction price under any transaction condition is less than a seventh preset threshold, screening the N initial influencing factors again to obtain multiple target influencing factors.

[0012] According to another aspect of an embodiment of the present application, a device for determining an influencing factor is also provided, including: an acquisition module for acquiring target historical data of a target trading system, wherein the target historical data includes: historical evaluation prices, historical transaction prices, and historical transaction evaluation values; a determination module for determining N initial influencing factors based on the target historical data, wherein the initial influencing factors are factors that affect the transaction price of the target trading system, and N is an integer greater than or equal to 2; a screening module for screening the N initial influencing factors to obtain M target influencing factors, wherein the transaction price of the target trading system determined based on the M target influencing factors is greater than or equal to the transaction price of the target trading system determined based on any M initial influencing factors among the N initial influencing factors, and M is a positive integer less than or equal to N.

[0013] According to another aspect of the embodiments of the present application, a computer-readable storage medium is provided, in which a computer program is stored. The computer program is configured to execute the above-mentioned method for determining the influencing factor when running.

[0014] According to another aspect of an embodiment of the present application, an electronic device is provided, including a memory, a processor, and a computer program stored in the memory and executable on the processor, wherein the processor executes the method for determining the influencing factor through the computer program.

[0015] According to another aspect of the embodiments of the present application, a computer program product is provided, including a computer program, and when the computer program is executed by a processor, the method for determining the influencing factor is implemented.

[0016] Through this application, historical data of the trading system is obtained to determine the initial influencing factors that affect the trading price of the trading system. The initial influencing factors are then screened to obtain multiple influencing factors that have a greater impact on the trading price of the trading system. Since manual guessing of the influencing factors based on experience is avoided, the accuracy and efficiency of determining the influencing factors that affect the trading system are improved, thereby solving the problem of low accuracy and efficiency in determining the factors that affect the trading system. BRIEF DESCRIPTION OF THE DRAWINGS

[0017] The accompanying drawings, which are incorporated in and constitute a part of this specification, illustrate embodiments consistent with the present application and, together with the description, serve to explain the principles of the present application.

[0018] In order to more clearly illustrate the embodiments of the present application or the technical solutions in the prior art, the following briefly introduces the drawings required for use in the embodiments or the description of the prior art. Obviously, for ordinary technicians in this field, other drawings can be obtained based on these drawings without any creative work.

[0019] Figure 1 This is a hardware structure block diagram of a mobile terminal for a method for determining an influencing factor according to an embodiment of the present application;

[0020] Figure 2 is a flow chart of a method for determining an impact factor according to an embodiment of the present application;

[0021] Figure 3 This is a structural block diagram of a device for determining an impact factor according to an embodiment of the present application. DETAILED DESCRIPTION

[0022] In order to enable those skilled in the art to better understand the present invention, the following will clearly and completely describe the technical solutions in the embodiments of the present invention in conjunction with the drawings in the embodiments of the present invention. Obviously, the described embodiments are only part of the embodiments of the present invention, not all of the embodiments. Based on the embodiments in the present invention, all other embodiments obtained by ordinary technicians in this field without making creative efforts should fall within the scope of protection of this application.

[0023] It should be noted that the terms "first", "second", etc. in the specification and claims of the present application and the above-mentioned drawings are used to distinguish similar objects and are not necessarily used to describe a specific order or sequential order. It should be understood that the data used in this way can be interchangeable where appropriate, so that the embodiments of the present application described herein can be implemented in a sequence other than those illustrated or described herein. In addition, the terms "including" and "having" and any of their variations are intended to cover non-exclusive inclusions, for example, a process, method, system, product or device comprising a series of steps or units is not necessarily limited to those steps or units clearly listed, but may include other steps or units that are not clearly listed or inherent to these processes, methods, products or devices.

[0024] The method embodiments provided in the embodiments of the present application can be executed in a mobile terminal, a computer terminal or a similar computing device. Taking running on a mobile terminal as an example, Figure 1This is a hardware structure block diagram of a mobile terminal for determining an influencing factor according to an embodiment of the present application. Figure 1 As shown, the mobile terminal may include one or more ( Figure 1 The mobile terminal includes a processor 102 (only one of which is shown) (the processor 102 may include but is not limited to a microprocessor (Central Processing Unit, MCU) or a programmable logic device (Field Programmable Gate Array, FPGA) and a memory 104 for storing data. The mobile terminal may also include a transmission device 106 and an input / output device 108 for communication functions. It will be understood by those skilled in the art that Figure 1 The structure shown is only for illustration and does not limit the structure of the mobile terminal. Figure 1 More or fewer components than shown, or with Figure 1 Different configurations shown.

[0025] The memory 104 can be used to store computer programs, for example, software programs and modules of application software, such as the computer program corresponding to the method for determining the influencing factor in the embodiment of the present application. The processor 102 executes various functional applications and data processing by running the computer program stored in the memory 104, that is, implementing the above method. The memory 104 may include a high-speed random access memory, and may also include a non-volatile memory, such as one or more magnetic storage devices, flash memory, or other non-volatile solid-state memory. In some instances, the memory 104 may further include a memory remotely arranged relative to the processor 102, and these remote memories may be connected to the mobile terminal via a network. Examples of the above-mentioned network include but are not limited to the Internet, an intranet, a local area network, a mobile communication network, and a combination thereof.

[0026] Transmission device 106 is used to receive or transmit data via a network. A specific example of the aforementioned network may include a wireless network provided by the mobile terminal's communications provider. In one embodiment, transmission device 106 includes a network interface controller (NIC), which can be connected to other network devices via a base station to enable communication with the Internet. In another embodiment, transmission device 106 may be a radio frequency (RF) module, which is used to communicate with the Internet wirelessly.

[0027] In this embodiment, a method for determining an impact factor is provided. Figure 2 is a flow chart of a method for determining an impact factor according to an embodiment of the present application, such as Figure 2As shown, the process includes the following steps S202 to S206:

[0028] Step S202: Acquire target historical data of the target trading system, wherein the target historical data includes: historical evaluation prices, historical transaction prices, and historical transaction evaluation values;

[0029] Optionally, original historical data of the target trading system is obtained and preprocessed to obtain target historical data, including:

[0030] Identify and address outliers by calculating the median and median absolute deviation (MAD) of each column of data (historical appraisal price, historical transaction price, and historical transaction appraisal value). Set the upper and lower thresholds for outliers to the median plus or minus three times the MAD. This way, only values ​​that deviate significantly from the central trend are identified as outliers and replaced with the corresponding thresholds, minimizing the impact of outliers on the overall analysis.

[0031] For missing values ​​(null values) in the original data, we adopt the industry mean filling strategy. That is, we use the average value of the non-null sample data in the same industry to fill these null values. This not only maintains the integrity of the data, but also utilizes the characteristics of the industry, improving the comparability of the data and the effectiveness of the analysis.

[0032] Integrate indicator values ​​and merge raw data indicators that reflect similar characteristics or have internal connections to reduce data redundancy, while enhancing the logic of factor construction and data consistency, facilitating subsequent factor calculation and analysis;

[0033] Finally, all raw indicator values ​​were standardized by subtracting the mean from each indicator value and then dividing by the standard deviation. This process transformed all data onto a common scale, eliminating the effects of dimension and ensuring that different indicators could be compared and weighted fairly in multifactor analysis.

[0034] Step S204: determining N initial impact factors based on the target historical data, wherein the initial impact factors are factors that affect the transaction price of the target trading system, and N is an integer greater than or equal to 2;

[0035] Optionally, in order to improve the efficiency of determining the initial influencing factors, a computing engine is used to determine N initial influencing factors based on the target historical data, wherein the computing engine adopts a distributed processing method, and the built-in influencing factor library contains a variety of factors designed for market characteristics, such as commodity attributes, market trends, seasonal factors, trader behavior and other factors, including but not limited to curve level factors, curve slope factors, credit spread factors, rating spread factors, carry factors, quality factors, momentum factors, scale factors, liquidity factors. These factors cover multiple dimensions of the trading system and can capture various factors that may affect trading prices from macro to micro levels.

[0036] Optionally, when determining the initial impact factors, structured data is selected to avoid the complexity and uncertainty that unstructured data may bring, thereby improving the efficiency and accuracy of data processing. Specifically, from the perspective of underlying data, impact factors directly related to the transaction type are used. These structured underlying factors are then input into the calculation engine, which dynamically configures and calculates the initial impact factors. The calculation engine combines target historical data from the trading market and utilizes advanced parallel computing technology to significantly reduce the time required for factor calculation, making large-scale factor analysis possible.

[0037] Step S206: Screen the N initial influencing factors to obtain M target influencing factors, wherein the transaction price of the target trading system determined based on the M target influencing factors is greater than or equal to the transaction price of the target trading system determined based on any M initial influencing factors among the N initial influencing factors, and M is a positive integer less than or equal to N.

[0038] It should be noted that when using M target influencing factors to predict transaction prices, the prediction accuracy and stability exceed those achieved using any other initial factor set of the same number (M). This demonstrates that the M target influencing factors are optimally selected and can more accurately reflect the core drivers of transaction price fluctuations.

[0039] It should be noted that steps S202 to S206 effectively identify and screen the factors most influential on price fluctuations in the trading system. Compared to traditional methods of selecting factors based on intuition or experience, quantitative analysis and efficient calculations significantly improve the accuracy and efficiency of factor identification, reduce forecasting errors caused by subjective judgment, and adapt to the needs of real-time analysis of large-scale data. This method provides strong support for trading system strategy optimization, risk management, and price forecasting, helping to enhance the scientific nature of trading decisions and market competitiveness.

[0040] The above steps obtain the historical data of the trading system, thereby determining the initial influencing factors that affect the trading price of the trading system, and then screen the initial influencing factors to obtain multiple influencing factors that have a greater impact on the trading price of the trading system. Since manual guessing of the influencing factors based on experience is avoided, the accuracy and efficiency of determining the influencing factors that affect the trading system are improved, thereby solving the problem of low accuracy and efficiency in determining the factors that affect the trading system.

[0041] In an exemplary embodiment, screening the N initial impact factors to obtain M target impact factors can be achieved by following the steps S11-S13:

[0042] Step S11: Determine N information coefficients and N information coefficient ratios based on the N initial influencing factors, wherein the i-th information coefficient among the N information coefficients is the information coefficient between the i-th initial influencing factor among the N initial influencing factors and the historical transaction price, and the i-th information coefficient ratio among the N information coefficient ratios is the information coefficient ratio between the i-th initial influencing factor and the historical transaction price. The information coefficient is used to measure the degree of correlation between the influencing factor and the historical transaction price, and the information coefficient ratio is used to measure the stability of the correlation. i is a positive integer less than or equal to N.

[0043] Optionally, in this preliminary screening step, the correlation and stability between each initial influencing factor (a total of N) and the historical transaction price are first measured by statistical calculation. Specifically, for each initial influencing factor (denoted as the i-th factor), the influencing factor determination system will calculate the information coefficient (IC) between it and the historical transaction price. This step is essentially to evaluate the strength of the linear relationship between the factor and the price trend. Subsequently, the information coefficient ratio (ICR) is calculated, which usually involves the time series standard deviation of the IC to evaluate the stability of the correlation. IC and ICR are both key indicators for measuring the predictive ability of influencing factors. The former focuses on immediate correlation, while the latter focuses more on the persistence of correlation.

[0044] Step S12: determining P candidate influencing factors from the N initial influencing factors according to the N information coefficients and the N information coefficient ratios, wherein the information coefficient of each candidate influencing factor of the P candidate influencing factors is greater than a first preset threshold, and the information coefficient ratio is greater than a second preset threshold, and P is a positive integer less than or equal to N;

[0045] Optionally, a first preset threshold is used to filter the information coefficient. Only factors with an IC above this threshold are included in the candidate list, indicating that the candidate factors have a strong correlation with historical trading prices. A second preset threshold is used to filter the information coefficient ratio. Similarly, only factors with an ICR above this threshold are retained, ensuring the stability of the factors' price predictions. Therefore, the P candidate factors are a set of factors that are not only highly correlated with historical trading prices, but also have a fairly stable correlation.

[0046] Step S13: Determine the M target impact factors from the P candidate impact factors.

[0047] It should be noted that through the above steps, it is possible to accurately identify the set of factors that have a significant impact on trading prices from a large number of initial factors. This not only improves the accuracy of the prediction, but also ensures the simplicity and generalization ability of the determination process, providing strong support for the trading system.

[0048] In an exemplary embodiment, screening the N initial impact factors to obtain M target impact factors can be achieved by following the steps S21-S25:

[0049] Step S21: dividing the N initial impact factors into X groups of impact factors according to a first preset classification condition, wherein each group of impact factors in the X groups of impact factors includes at least one initial impact factor, and X is a positive integer less than or equal to N;

[0050] Optionally, the first preset classification condition may be the attributes and sources of the influencing factors, for example, the N initial influencing factors are divided into different groups such as macroeconomic factors, industry-specific factors, and transaction characteristic factors.

[0051] Step S22: Divide each of the X groups of impact factors into a plurality of subgroups according to a second preset classification condition, to obtain Y subgroups, wherein each of the Y subgroups includes at least one initial impact factor, and Y is an integer greater than or equal to X;

[0052] Optionally, the influencing factors within each group are further subdivided according to the second preset classification criteria to form Y subgroups. This process involves further analysis of the factors' attributes (the second preset classification criteria), such as time series characteristics and volatility characteristics, or by cluster analysis of the inter-factor covariance matrix to group highly correlated or synergistic factors.

[0053] Step S23: calculating the transaction price corresponding to each of the Y groups based on the initial impact factor of each group;

[0054] Optionally, the initial impact factors within each group are modeled with historical transaction prices to calculate the expected transaction price under the group's initial impact factor configuration. This can be done through regression analysis, machine learning models or other statistical methods, aiming to evaluate the explanatory power and predictive accuracy of different factor combinations on transaction prices.

[0055] Step S24: using the initial impact factors of one or more target groups as candidate impact factors to obtain Z candidate impact factors, wherein the transaction price corresponding to the target group is greater than a third preset threshold, and Z is a positive integer less than or equal to N;

[0056] Optionally, groups whose trading prices exceed a third preset threshold are considered target groups, and all initial influencing factors extracted from these groups constitute Z candidate influencing factors. The third preset threshold here can be excess returns compared to the market benchmark, returns within the forecast error range, or any other standard representing the impact of the influencing factors, to ensure that the candidate factor set has significant market explanatory power or predictive advantage.

[0057] Optionally, after step S24, more detailed grouping may be performed, ie, multi-layer and multi-level screening, to obtain a certain number of candidate influencing factors.

[0058] Step S25: Determine the M target impact factors from the Z candidate impact factors.

[0059] It's important to note that the above steps, through multi-level classification and group screening, systematically evaluate and compare the impact of various influencing factors, avoiding the bias of unilateral reliance on a single indicator or method, and enhancing the objectivity and comprehensiveness of the screening results. The transaction prices calculated based on the group can reflect the unique ability of different factor combinations to predict market behavior. Through comparison and selection, the final M target influencing factors are more likely to capture the true dynamics of the market and improve forecast accuracy.

[0060] In an exemplary embodiment, when W is equal to P or Z, determining the M target influencing factors from the W candidate influencing factors can be achieved by the following steps: using forward step screening method and / or backward stepwise elimination method and / or stability selection method to determine the M target influencing factors from the W candidate influencing factors.

[0061] It should be noted that the forward step screening method is an incremental model-building approach. The initial model contains only the intercept term. Then, candidate influencing factors with the strongest correlation with the target variable (trading price) are gradually introduced until the model's explanatory power (usually measured by the adjusted coefficient of determination (R²)) increases, falling below a pre-set threshold (e.g., ΔR² < 0.01) or reaching the maximum number of factors. This process helps identify the factors with the greatest explanatory power for trading prices.

[0062] Alternatively, assume there are 10 candidate impact factors (W = 10). Using the forward step screening method, first build a model containing only the intercept term. Then, gradually add candidate impact factors to the model based on the Pearson correlation coefficient of each candidate impact factor until the explanatory power of the model is no longer sufficient to introduce new candidate impact factors (e.g., ΔR² < 0.01). At this point, the candidate impact factors in the model are the M target impact factors.

[0063] It should be noted that the backward stepwise elimination method starts with a full model containing all candidate influencing factors and gradually removes those candidate influencing factors with the smallest contribution to the model (usually non-significant factors with the largest p-value (a metric used in hypothesis testing in statistics), such as p>0.05) until the elimination of any candidate influencing factor causes the model's Akaike Information Criterion (AIC) value to increase by more than a preset critical value (such as ΔAIC>2) or reaches the minimum number of factors required. This method helps eliminate redundant factors, further simplify the model, and improve predictive accuracy.

[0064] Alternatively, assuming there are 15 candidate influencing factors (W=15), a model can be constructed that includes all 15 candidate influencing factors. Then, the candidate influencing factors with the largest p-values ​​are removed one by one from the model until the AIC value of the model increases significantly or the effective factor number limit is reached (for example, only 5 candidate influencing factors are retained). The remaining candidate influencing factors are the M target influencing factors.

[0065] It should be noted that the stability selection method uses multiple random samplings to traverse all possible factor subsets and perform Least Absolute Shrinkage and Selection Operator Regression (LASSO regression) on each subset. LASSO regression is a regularized regression method that can simultaneously perform variable selection and parameter estimation. The frequency of each factor being selected is counted, and those factors with frequencies exceeding a percentage threshold (for example, greater than 70%) are selected. Variables with excessively high Variance Inflation Factors (VIFs), a measure of collinearity, are removed to construct the final factor combination. This method can identify factors that exhibit stable predictive ability under different sampling conditions, improving the robustness and generalization ability of the model.

[0066] Alternatively, assuming there are 20 candidate influencing factors (W = 20), the stability selection method randomly selects a subset of these 20 factors, for example, 10 factors at a time, for LASSO regression. This process is repeated multiple times (for example, 100 times), and the number of times each factor is selected is counted. If a factor is selected in at least 70% of the regressions and its VIF is below 5, it is considered a stable and independently contributing factor and is added to the target factor set.

[0067] It should be noted that through the above steps, not only can the quality of the selected influencing factors be ensured, but also a refined and powerful model can be constructed, providing solid data support and decision-making basis for the management and optimization of the trading system.

[0068] In an exemplary embodiment, after obtaining M target impact factors, the method further includes the following steps S31-S34:

[0069] Step S31: Divide the preset time period into E sub-time periods, where E is a positive integer;

[0070] Step S32: Determine the transaction price of the target trading system in each sub-time period of the E sub-time periods to obtain E transaction prices, wherein the transaction price of the e-th sub-time period of the E sub-time periods is determined based on the M target impact factors and the weights corresponding to the M target impact factors, each of the M target impact factors has a different weight corresponding to different sub-time periods, and e is a positive integer less than or equal to E;

[0071] Alternatively, if the preset time period is one year, it can be divided into four sub-periods, i.e., four quarters, based on quarterly factors. For each quarter, the weights corresponding to the M target influencing factors are determined based on the current market environment. For example, if there are three target influencing factors (M=3), namely factor a, factor b, and factor c, then in the first quarter, the weight corresponding to factor a is 0.2, the weight corresponding to factor b is 0.3, and the weight corresponding to factor c is 0.5; in the second quarter, the weight corresponding to factor a is 0.3, the weight corresponding to factor b is 0.4, and the weight corresponding to factor c is 0.3; in the third quarter, the weight corresponding to factor a is 0.4, the weight corresponding to factor b is 0.5, and the weight corresponding to factor c is 0.1; and in the fourth quarter, the weight corresponding to factor a is 0.5, the weight corresponding to factor b is 0.1, and the weight corresponding to factor c is 0.4. The transaction price corresponding to each sub-period is calculated.

[0072] Step S33: Calculating target indicator values ​​corresponding to the target trading system within the preset time period based on the E transaction prices, wherein the target indicator values ​​include: rate of return, Sharpe ratio, and maximum drawdown value;

[0073] It should be noted that the Sharpe ratio is used to measure the excess return that a trading portfolio can obtain for every unit of total risk it bears. A higher Sharpe ratio means that the portfolio has a higher excess return for the same risk, or that the risk taken is lower for the same excess return.

[0074] It's important to note that Maximum Drawdown (MDD) is a key metric for measuring the risk of a trading portfolio or strategy, particularly important in assessing losses and strategy robustness. It's defined as the magnitude of the loss from the highest point in an asset's value to the subsequent lowest point over a selected time period, typically expressed as a percentage.

[0075] Step S34: When the rate of return is less than the fourth preset threshold, and / or the Sharpe ratio is less than the fifth preset threshold, and / or the maximum drawdown value is greater than the sixth preset threshold, the N initial influencing factors are screened again to obtain multiple target influencing factors.

[0076] Optionally, if the return over the entire preset time period is less than 2%, the Sharpe ratio is less than 0.5, or the maximum drawdown value exceeds -10%, all N initial influencing factors are reviewed and the target influencing factor portfolio is reconstructed.

[0077] It's important to note that in the above steps, factor weights change with market conditions, ensuring more informed investment decisions and preventing the need to cling to outdated strategies due to shifting market conditions. Real-time calculation of key performance indicators such as return, Sharpe ratio, and maximum drawdown helps promptly identify and adjust investment strategy deficiencies, mitigating investment risk. When a strategy underperforms, proactively re-screening the influencing factors can be considered a self-learning and optimization process for the strategy, continuously seeking more effective combinations of factors and weights to enhance long-term returns.

[0078] In an exemplary embodiment, after obtaining M target impact factors, the method further includes the following steps S41-S43:

[0079] Step S41: adjusting the transaction conditions of the target transaction system, wherein the transaction conditions include: transaction goods type and transaction duration;

[0080] It should be noted that adjusting trading conditions means changing certain parameters or preferences in the trading strategy based on changes in the market environment and the needs of trading goals.

[0081] Step S42: determining the transaction price of the target trading system under different transaction conditions based on the M target impact factors;

[0082] Optionally, after determining the M target influencing factors, these factors and their weights are further used to calculate the reasonable price or expected return of the trading system under specific trading conditions.

[0083] Step S43: When the transaction price under any transaction condition is less than the seventh preset threshold, the N initial impact factors are screened again to obtain a plurality of target impact factors.

[0084] Alternatively, if the calculated transaction price under certain trading conditions falls below a preset threshold (the seventh preset threshold), this may indicate that the current trading strategy and target influencing factor combination are not suitable for the market environment under these conditions. Economic forecasts based on the current target influencing factors may face a significant risk of loss or fail to achieve the expected return level. All initial influencing factors (N) are then reviewed to select the influencing factors that best meet the current market conditions.

[0085] It’s important to note that this dynamic adjustment mechanism enables trading strategies to better adapt to market changes, improving their overall adaptability and flexibility. Continuously optimizing the influencing factor combination allows for a more precise match between the trading system’s characteristics and market trends.

[0086] Obviously, the embodiments described above are only part of the embodiments of the present invention, rather than all the embodiments. In order to better understand the above method, the above process is described below in conjunction with the embodiments, but it is not intended to limit the technical solutions of the embodiments of the present invention. Specifically:

[0087] This application provides a multi-factor analysis method for trading systems, enabling multi-dimensional data integration, dynamic factor construction, weight optimization, multi-factor analysis, strategy execution monitoring, and cross-sample validation. It establishes a factor calculation engine, standardizes factor construction, and improves targeting. The use of structured data and parallel computing significantly improves computational efficiency. The factor screening process constructs a multi-level quantitative standard validation method and system to enhance forecast accuracy.

[0088] The embodiment of the present application includes the following modules: Data processing module: obtains raw data and pre-processes the raw data; Factor construction module: based on the processed data and the factor calculation engine, the factor construction is dynamically configured based on the trading market; Factor analysis module: obtains the factor effectiveness analysis results through single factor analysis and multi-factor analysis methods; Backtest execution module: dynamically synthesizes factors, sets the position adjustment cycle, and outputs the backtest results based on the factor construction; Cross-sample verification module: sets other sample spaces in the module to verify the universality of the factors. The specific description of each module is as follows:

[0089] Data Processing Module: This module obtains raw market transaction data from authoritative data sources and preprocesses the raw data. Data preprocessing includes the following steps: First, outlier processing is performed by calculating the median and median absolute deviation (MAD). Upper and lower thresholds are determined as the median plus or minus three times the MAD. Outliers exceeding these thresholds are replaced with the upper and lower thresholds. Second, null values ​​are processed. Null values ​​are filled with the average of non-null samples from the same industry. Third, raw indicator values ​​are merged to combine related indicator values. Fourth, raw indicator values ​​are standardized by subtracting the average indicator value from the value and dividing by the standard deviation.

[0090] The Factor Construction Module inputs raw data into the calculation engine and outputs factor results based on market dynamics. Factor calculations utilize a distributed approach to improve efficiency. The built-in factor library includes curve level factors, curve slope factors, quality factors, momentum factors, scale factors, liquidity factors, and custom factors. The use of structured data avoids the use of unstructured data, systematically improving calculation efficiency.

[0091] Factor Analysis Module: Constructs multi-factor analysis models, conducts process linkage design, and tests factor effectiveness, including:

[0092] Step 1, calculate IC and ICR values;

[0093] In step 1.1, calculate the standardized value of the impact factor in period t (predetermined sub-period) and the holding return of the trading system in period t+1, using the original data. Calculate the rank correlation coefficient of the two columns as the IC value. In step 1.2, repeat step 1.1 to calculate the IC value for all t periods. In step 1.3, calculate the ratio of the mean to the standard deviation of the IC values ​​for multiple periods as the ICR value. In step 1.4, calculate the t-statistic to determine if the mean of the IC values ​​for multiple periods is significantly different from zero. In step 1.5, set thresholds for the IC value, ICR value, and t-statistic. Factors exceeding the threshold will proceed to the next step of testing.

[0094] Step 2, calculating the layered test results;

[0095] Step 2.1: Divide the factors into n layers in period t and perform weighting within each layer. Step 2.2: Repeat step 2.1 to calculate the impact factors for all periods t. Step 2.3: Based on the impact factors for all periods t in step 2.2, calculate the annualized rate of return, Sharpe ratio, and maximum drawdown for all layers. Test the monotonicity of the indicator results between layers. Step 2.4: Use conditional double sorting to construct the layered portfolio for period t. is the control variable (attribute of the influencing factor), is the factor to be tested. The conditional double sorting is first used The ranking divides the impact factors into groups (e.g. ), and then use The ranking is further divided into groups (e.g. ). After being controlled To determine whether there is incremental information about the transaction price, we only need to focus on Construct factors and calculate factor returns. In group indivual Groups with the same ranking are combined to form Combinations are hierarchical combinations. Taking combination 1 as an example, let the final hierarchical combination be ,but It is obtained by the following union: , and so on for other combinations. Step 2.5: Repeat Step 2.4 to calculate the trading prices for all t periods based on conditional double or multiple ranking. Step 2.6: Based on the trading prices for all t periods in Step 2.5, calculate the annualized rate of return, Sharpe ratio, and maximum drawdown for all strata. Test the monotonicity of the indicator results between strata. Step 2.7: Influencing factors that pass the monotonicity tests in Steps 2.3 and 2.6 proceed to the next step of testing.

[0096] Step 3: Set up the Fama-MacBeth multi-factor regression subsystem;

[0097] Build a regression model for factors that pass the tests in steps 1 and 2. Screen multi-factor combinations using one of the following methods: Method 1: Forward stepwise screening. Construct a baseline model containing only the intercept term. Based on the Pearson correlation coefficients between factors and the target variable, introduce the factors with the strongest correlations one by one until the improvement in model goodness-of-fit (R²) falls below a preset threshold (e.g., ΔR² < 0.01) or the maximum number of factors is reached. Method 2: Backward stepwise elimination. Initialize the full factorial regression model and sequentially remove non-significant factors with the highest t-test p-values ​​(e.g., p > 0.05) until removing any factor increases the model AIC by more than a preset threshold (e.g., ΔAIC > 2) or the minimum number of factors is reached. Method 3: Stability selection. Iterate through all factor subsets and perform LASSO regression on each subset. Count the frequencies of factors selected and select factors with frequencies exceeding a threshold (e.g., greater than 70%) to form the final combination. Collinear variables with high variance inflation factors (e.g., VIF > 5) are also removed.

[0098] Backtest Execution Module: Step 1: Dynamically synthesize factors using the IC or ICR values ​​of factors from the Factor Analysis Module as weights. The module supports dynamic adaptive optimization, with the system automatically detecting continuous changes in factor IC values ​​and adjusting the weights of each factor. Step 2: The module receives the input backtest period, the factors that have passed the test, and the preset period, calculates the transaction price, selects the factors with the highest performance ranking (e.g., the top 20%), sets weights, and incorporates a real-time monitoring mechanism, including maximum drawdown (with an early warning line) and factor exposure deviation. If risk control thresholds are triggered, the system automatically rebalances. Step 3: Calculate backtest results, including the cumulative yield curve and performance evaluation metrics such as annualized yield, Sharpe ratio, maximum drawdown, and Karma ratio.

[0099] Cross-sample Validation: To test the universality of factors and strategies, this module includes multiple, differentiated sample space testing systems. By dynamically adjusting factor screening thresholds and portfolio construction rules, it verifies the universality of strategies under different market environments and liquidity constraints. You can also set various screening criteria for sample switching.

[0100] It should be noted that the advantages of this application include: 1. It can accurately capture influencing factors and avoid the explanatory power limitations of traditional single factors; 2. It constructs a quantitative factor screening method and system, integrates multiple standardized methods to obtain results, reduces the negative impact of human interference on predictions, and improves the accuracy of factor predictions; 3. It realizes distributed calculations, reduces time consumption, and improves calculation efficiency.

[0101] Through the description of the above embodiments, those skilled in the art can clearly understand that the methods according to the above embodiments can be implemented by means of software plus the necessary general hardware platform. Of course, they can also be implemented by hardware, but in many cases the former is a better implementation method. Based on this understanding, the technical solution of the present application, or the part that contributes to the existing technology, can be embodied in the form of a software product. The computer software product is stored in a storage medium (such as ROM / RAM, magnetic disk, optical disk), and includes a number of instructions for enabling a terminal device (which can be a mobile phone, computer, server, or network device, etc.) to execute the methods of each embodiment of the present application.

[0102] In this embodiment, a device for determining an impact factor is also provided. The device for determining an impact factor is used to implement the above-mentioned embodiments and preferred embodiments, and the details already described will not be repeated. As used below, the term "module" can be a combination of software and / or hardware that implements a predetermined function. Although the devices described in the following embodiments are preferably implemented in software, implementation using hardware, or a combination of software and hardware, is also possible and contemplated.

[0103] Figure 3 : is a structural block diagram of a device for determining an impact factor according to an embodiment of the present application, the device comprising:

[0104] An acquisition module 302 is configured to acquire target historical data of a target trading system, wherein the target historical data includes: historical evaluation prices, historical transaction prices, and historical transaction evaluation values;

[0105] A determination module 304 is configured to determine N initial impact factors based on the target historical data, wherein the initial impact factors are factors that affect the transaction price of the target trading system, and N is an integer greater than or equal to 2;

[0106] The screening module 306 is used to screen the N initial influencing factors to obtain M target influencing factors, wherein the transaction price of the target trading system determined based on the M target influencing factors is greater than or equal to the transaction price of the target trading system determined based on any M initial influencing factors among the N initial influencing factors, and M is a positive integer less than or equal to N.

[0107] The above-mentioned device for determining the influencing factors obtains the historical data of the trading system, thereby determining the initial influencing factors that affect the trading price of the trading system, and then screens the initial influencing factors to obtain multiple influencing factors that have a greater impact on the trading price of the trading system. Since it avoids manual guessing of the influencing factors based on experience, it improves the accuracy and efficiency of determining the influencing factors that affect the trading system, thereby solving the problem of low accuracy and efficiency in determining the factors that affect the trading system.

[0108] In an exemplary embodiment, the screening module 306 is further configured to determine N information coefficients and N information coefficient ratios based on the N initial influencing factors, wherein the i-th information coefficient among the N information coefficients is the information coefficient between the i-th initial influencing factor among the N initial influencing factors and the historical transaction price, and the i-th information coefficient ratio among the N information coefficient ratios is the information coefficient ratio between the i-th initial influencing factor and the historical transaction price, the information coefficient is used to measure the degree of correlation between the influencing factor and the historical transaction price, and the information coefficient ratio is used to measure the stability of the correlation, and i is a positive integer less than or equal to N; determine P candidate influencing factors from the N initial influencing factors based on the N information coefficients and the N information coefficient ratios, wherein the information coefficient of each candidate influencing factor among the P candidate influencing factors is greater than a first preset threshold value, and the information coefficient ratio is greater than a second preset threshold value, and P is a positive integer less than or equal to N; and determine the M target influencing factors from the P candidate influencing factors.

[0109] In an exemplary embodiment, the screening module 306 is further configured to divide the N initial influencing factors into X groups of influencing factors according to a first preset classification condition, wherein each group of influencing factors in the X groups of influencing factors includes at least one initial influencing factor, and X is a positive integer less than or equal to N; divide each group of influencing factors in the X groups of influencing factors into a plurality of small groups according to a second preset classification condition, thereby obtaining Y small groups, wherein each of the Y small groups includes at least one initial influencing factor, and Y is an integer greater than or equal to X; calculate a transaction price corresponding to each of the Y small groups based on the initial influencing factors in each of the Y small groups; use the initial influencing factors in one or more target groups as candidate influencing factors, thereby obtaining Z candidate influencing factors, wherein the transaction price corresponding to the target group is greater than a third preset threshold, and Z is a positive integer less than or equal to N; and determine the M target influencing factors from the Z candidate influencing factors.

[0110] In an exemplary embodiment, the screening module 306 is further used to determine the M target influencing factors from the W candidate influencing factors by using a forward step screening method and / or a backward stepwise elimination method and / or a stability selection method when W is equal to P or Z.

[0111] In an exemplary embodiment, the apparatus further includes: a verification module configured to, after obtaining M target impact factors, divide a preset time period into E sub-time periods, where E is a positive integer; determine a transaction price of the target trading system in each of the E sub-time periods to obtain E transaction prices, wherein the transaction price of the e-th sub-time period in the E sub-time periods is determined based on the M target impact factors and their respective corresponding weights, each of the M target impact factors having a different corresponding weight in different sub-time periods, and e being a positive integer less than or equal to E; calculate a target indicator value corresponding to the target trading system in the preset time period based on the E transaction prices, wherein the target indicator value includes: rate of return, Sharpe ratio, and maximum drawdown value; and a screening module 306 configured to, when the rate of return is less than a fourth preset threshold, and / or the Sharpe ratio is less than a fifth preset threshold, and / or the maximum drawdown value is greater than a sixth preset threshold, screen the N initial impact factors again to obtain multiple target impact factors.

[0112] In an exemplary embodiment, the verification module is further used to adjust the transaction conditions of the target trading system after obtaining M target influencing factors, wherein the transaction conditions include: the type of trading goods and the transaction duration; and determine the transaction price of the target trading system under different transaction conditions based on the M target influencing factors; the screening module 306 is further used to screen the N initial influencing factors again to obtain multiple target influencing factors when the transaction price under any transaction condition is less than a seventh preset threshold.

[0113] An embodiment of the present application further provides a computer-readable storage medium, in which a computer program is stored. The computer program is configured to execute the steps of any of the above method embodiments when run.

[0114] Optionally, in this embodiment, the storage medium may be configured to store a computer program for performing the following steps:

[0115] S1, obtaining target historical data of a target trading system, wherein the target historical data includes: historical evaluation prices, historical transaction prices, and historical transaction evaluation values;

[0116] S2, determining N initial impact factors based on the target historical data, wherein the initial impact factors are factors that affect the transaction price of the target trading system, and N is an integer greater than or equal to 2;

[0117] S3. Screen the N initial influencing factors to obtain M target influencing factors, wherein the transaction price of the target trading system determined based on the M target influencing factors is greater than or equal to the transaction price of the target trading system determined based on any M initial influencing factors among the N initial influencing factors, and M is a positive integer less than or equal to N.

[0118] In an exemplary embodiment, the computer-readable storage medium may include, but is not limited to, various media that can store computer programs, such as a USB flash drive, a read-only memory (ROM), a random access memory (RAM), a mobile hard disk, a magnetic disk, or an optical disk.

[0119] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail here.

[0120] An embodiment of the present application further provides a computer program product, including a computer program, and the computer program performs the steps of any of the above method embodiments when executed by a processor.

[0121] An embodiment of the present application further provides another computer program product, comprising a non-volatile computer-readable storage medium, wherein the non-volatile computer-readable storage medium stores a computer program, and when the computer program is executed by a processor, the steps of any of the above method embodiments are implemented.

[0122] For specific examples in this embodiment, reference may be made to the examples described in the above embodiments and exemplary implementation modes, and this embodiment will not be described in detail here.

[0123] Obviously, those skilled in the art should understand that the modules or steps of the present application described above can be implemented using a general-purpose computing device, they can be concentrated on a single computing device, or distributed across a network composed of multiple computing devices, they can be implemented using program code executable by the computing device, and thus, they can be stored in a storage device and executed by the computing device, and in some cases, the steps shown or described can be performed in a different order than herein, or they can be fabricated into separate integrated circuit modules, or multiple modules or steps can be fabricated into a single integrated circuit module for implementation. Thus, the present application is not limited to any specific combination of hardware and software.

[0124] The above is only a preferred embodiment of the present application. It should be pointed out that for ordinary technicians in this technical field, several improvements and modifications can be made without departing from the principles of the present application. These improvements and modifications should also be regarded as the scope of protection of the present application.

Claims

1. A method for determining an impact factor, characterized in that: include: Acquiring target historical data of a target trading system, wherein the target historical data includes: historical evaluation prices, historical transaction prices, and historical transaction evaluation values; Determining N initial impact factors based on the target historical data, wherein the initial impact factors are factors that affect the transaction price of the target trading system, and N is an integer greater than or equal to 2; The N initial influencing factors are screened to obtain M target influencing factors, wherein the transaction price of the target trading system determined based on the M target influencing factors is greater than or equal to the transaction price of the target trading system determined based on any M initial influencing factors among the N initial influencing factors, and M is a positive integer less than or equal to N.

2. The method according to claim 1, characterized in that The N initial impact factors are screened to obtain M target impact factors, including: Determining N information coefficients and N information coefficient ratios based on N initial influencing factors, wherein the i-th information coefficient among the N information coefficients is the information coefficient between the i-th initial influencing factor among the N initial influencing factors and the historical transaction price, and the i-th information coefficient ratio among the N information coefficient ratios is the information coefficient ratio between the i-th initial influencing factor and the historical transaction price, the information coefficient is used to measure the degree of correlation between the influencing factor and the historical transaction price, and the information coefficient ratio is used to measure the stability of the degree of correlation, and i is a positive integer less than or equal to N; Determining P candidate influencing factors from the N initial influencing factors according to the N information coefficients and the N information coefficient ratios, wherein the information coefficient of each candidate influencing factor in the P candidate influencing factors is greater than a first preset threshold, and the information coefficient ratio is greater than a second preset threshold, and P is a positive integer less than or equal to N; The M target impact factors are determined from the P candidate impact factors.

3. The method according to claim 1, characterized in that The N initial impact factors are screened to obtain M target impact factors, including: Dividing the N initial impact factors into X groups of impact factors according to a first preset classification condition, wherein each group of impact factors in the X groups of impact factors includes at least one initial impact factor, and X is a positive integer less than or equal to N; Dividing each of the X groups of impact factors into a plurality of subgroups according to a second preset classification condition to obtain Y subgroups, wherein each of the Y subgroups includes at least one initial impact factor, and Y is an integer greater than or equal to X; Calculating the transaction price corresponding to each of the Y groups based on the initial impact factor of each group; Taking the initial impact factors of one or more target groups as candidate impact factors, and obtaining Z candidate impact factors, wherein the transaction price corresponding to the target group is greater than a third preset threshold, and Z is a positive integer less than or equal to N; The M target impact factors are determined from the Z candidate impact factors.

4. The method according to claim 2 or 3, characterized in that When W is equal to P or Z, determining the M target impact factors from the W candidate impact factors includes: The M target influencing factors are determined from the W candidate influencing factors by adopting a forward step screening method and / or a backward stepwise elimination method and / or a stability selection method.

5. The method according to claim 1, wherein After obtaining the M target impact factors, the method further includes: Divide the preset time period into E sub-time periods, where E is a positive integer; Determining a transaction price of the target trading system in each sub-time period of the E sub-time periods to obtain E transaction prices, wherein the transaction price of the e-th sub-time period of the E sub-time periods is determined based on the M target impact factors and weights corresponding to the M target impact factors, each of the M target impact factors has a different weight corresponding to different sub-time periods, and e is a positive integer less than or equal to E; Calculating target indicator values ​​corresponding to the target trading system within the preset time period based on the E transaction prices, wherein the target indicator values ​​include: rate of return, Sharpe ratio, and maximum drawdown value; When the rate of return is less than the fourth preset threshold, and / or the Sharpe ratio is less than the fifth preset threshold, and / or the maximum drawdown value is greater than the sixth preset threshold, the N initial influencing factors are screened again to obtain multiple target influencing factors.

6. The method according to claim 1, characterized in that After obtaining the M target impact factors, the method further includes: Adjusting the transaction conditions of the target transaction system, wherein the transaction conditions include: transaction goods type and transaction duration; Determining a transaction price of the target trading system under different transaction conditions based on the M target influencing factors; When the transaction price under any transaction condition is less than the seventh preset threshold, the N initial impact factors are screened again to obtain a plurality of target impact factors.

7. A device for determining an impact factor, characterized in that: include: An acquisition module, configured to acquire target historical data of a target trading system, wherein the target historical data includes: historical evaluation prices, historical transaction prices, and historical transaction evaluation values; a determination module, configured to determine N initial impact factors based on the target historical data, wherein the initial impact factors are factors that affect the transaction price of the target trading system, and N is an integer greater than or equal to 2; A screening module is used to screen the N initial influencing factors to obtain M target influencing factors, wherein the transaction price of the target trading system determined based on the M target influencing factors is greater than or equal to the transaction price of the target trading system determined based on any M initial influencing factors among the N initial influencing factors, and M is a positive integer less than or equal to N.

8. A computer-readable storage medium, characterized in that: The computer-readable storage medium includes a stored program, wherein the program executes the method according to any one of claims 1 to 6 when executed.

9. An electronic device comprising a memory and a processor, characterized in that: A computer program is stored in the memory, and the processor is configured to execute the method according to any one of claims 1 to 6 through the computer program.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the method according to any one of claims 1 to 6 is implemented.

Images