Domain division-based index selection strategy construction system, method and equipment, medium and program product

Abstract

The invention provides a domain division-based index selection strategy construction system, method and equipment, a medium and a program product, which can be applied to the field of time sequence data analysis and financial science and technology. The system comprises a data import module which responds to an obtained first selection index, calls a first data interface based on the first selection index to determine a second time sequence data set in a first time interval from a first time sequence data set, and imports the second time sequence data set; the index calculation module is used for carrying out index calculation on the second time sequence data set based on the first selection index and a second index set divided in different categories from the first selection index to obtain a first index calculation result; the index analysis module performs index analysis calculation according to the first index calculation result to obtain a first index analysis result; and the domain analysis module is used for determining a second selection index associated with the first selection index from the second index set based on the first index analysis result and the second time sequence data set. According to the embodiment of the invention, the time sequence data screening efficiency can be improved.

Description

Technical Field

[0001] The present disclosure relates to the fields of time series data analysis and financial technology, and more specifically to a system, method, device, medium and program product for constructing an indicator selection strategy based on domain division. Background Art

[0002] Time series data, as an essential component of the data landscape, is widely present in many fields, including finance, industry, and meteorology. It is a sequence of observations arranged in chronological order, and its core characteristic is the temporal dependence between data points. This temporal dependence means that data at a given moment does not exist in isolation but is interconnected and influenced by past and future data. For example, in the stock market, today's stock price trends are influenced by data such as yesterday's closing price and trading volume, which in turn influence tomorrow's stock price. In industrial production, equipment operating parameters at different times also have a close temporal dependence, with parameters such as temperature and pressure at one moment affecting the equipment's operating status at a later moment.

[0003] Furthermore, time series data often has significant characteristics such as large data volumes and complex indicator relationships. With the continuous advancement of technology and the increasing diversity of data collection methods, the speed of data generation is accelerating, and the amount of data is also growing explosively.

[0004] Given these characteristics of time series data, filtering and processing this type of data requires computers with powerful storage and processing capabilities to handle such massive amounts of data. Traditional storage devices and computing methods often struggle with large-scale time series data, leading to problems such as slow processing speeds and insufficient storage space.

[0005] In actual time series data processing, even seemingly simple policy requirements from users trigger extensive computation. This is especially true for tasks involving full-data comparative analysis and multi-parameter analysis. This undoubtedly increases computational complexity and workload, placing higher demands on computer computing power and processing efficiency. In existing technologies, time series data users can only request simple data selection. Summary of the Invention

[0006] In view of the above problems, the present disclosure provides a domain-based indicator selection strategy construction system, method, device, medium and program product for improving the efficiency of time series data screening.

[0007] According to the first aspect of the present disclosure, a domain-based indicator selection strategy construction system is provided, comprising: a data import module, configured to, in response to obtaining a first selection indicator from a client, call a first data interface based on the first selection indicator to determine and import a second time series data set within a first time interval from a first time series data set. An indicator calculation module, configured to perform indicator calculation on the second time series data set based on the first selection indicator and a second indicator set divided into different categories from the first selection indicator, to obtain a first indicator calculation result. An indicator analysis module, configured to perform indicator analysis and calculation based on the first indicator calculation result, to obtain a first indicator analysis result. A domain analysis module, configured to determine a second selection indicator associated with the first selection indicator from the second indicator set based on the first indicator analysis result and the second time series data set.

[0008] According to an embodiment of the present disclosure, the system further includes: a second data interface, configured to crawl public data information related to individuals in the second time series data set within a first time interval from public data according to a call; an information analysis module, configured to determine the public data analysis results of the individuals in the second time series data set based on the public data information. The indicator analysis module is further configured to perform indicator analysis calculations based on the public data analysis results to obtain second indicator analysis results. The domain analysis module is further configured to determine a third selection indicator associated with the first selection indicator based on the second indicator analysis results.

[0009] According to an embodiment of the present disclosure, the indicator calculation module is further configured to perform indicator analysis and calculation on the first time series dataset based on a fourth selection indicator set associated with the second selection indicator to obtain a third indicator analysis and calculation result. The domain analysis module is further configured to determine, based on the third indicator analysis and calculation result, a fourth selection indicator associated with the first selection indicator from the fourth selection indicator set.

[0010] According to an embodiment of the present disclosure, the system further includes: a strategy building module configured to determine individuals of the second time series data set based on the first selection index and the second selection index.

[0011] Another aspect of an embodiment of the present disclosure provides a method for constructing an indicator selection strategy based on domain division, including: in response to obtaining a first selection indicator from a client, a data import module calls a first data interface based on the first selection indicator to determine and import a second time series data set within a first time interval from a first time series data set; an indicator calculation module performs a first indicator calculation on the second time series data set based on the first selection indicator and a second indicator set divided into different categories of the first selection indicator to obtain a first indicator calculation result; the first indicator calculation result is imported into an indicator analysis module to perform indicator analysis calculation to obtain a first indicator analysis result; the domain analysis module determines, from the second indicator set based on the first indicator analysis result and the second time series data set, a second selection indicator associated with the first selection indicator.

[0012] According to an embodiment of the present disclosure, the method further includes: the strategy construction module determining the individuals of the second time series data set based on the first selection index and the second selection index.

[0013] Another aspect of an embodiment of the present disclosure provides an electronic device, comprising: one or more processors; and a memory for storing one or more programs, wherein when the one or more programs are executed by the one or more processors, the one or more processors execute the above method.

[0014] Another aspect of an embodiment of the present disclosure provides a computer-readable storage medium having a computer program stored thereon, which, when executed by a processor, causes the processor to perform the above method.

[0015] Another aspect of an embodiment of the present disclosure provides a computer program product, including a computer program, which implements the above method when executed by a processor.

[0016] One or more of the above embodiments have the following beneficial effects:

[0017] An embodiment of the present disclosure proposes a domain-based indicator selection strategy construction system. In response to obtaining a first selection indicator from a client, a data import module calls a first data interface based on the first selection indicator to determine a second time series data set within a first time interval from a first time series data set and imports the second time series data set into the system, thereby avoiding full time series data processing and improving the efficiency of time series data screening and processing.

[0018] The embodiments of the present disclosure determine a second selection indicator associated with the first selection indicator from a second indicator set based on the first indicator analysis results and a second time series data set. By determining the first and second selection indicators, more selection indicators can be used for data screening, further improving data screening processing efficiency.

[0019] The embodiment of the present disclosure calls a first data interface to determine a second time series data set from a first time series data set, and calls a second data interface to crawl public data information related to individuals in the second time series data set within a first time interval from public data; this embodiment obtains data information from different data sources by calling different data interfaces, and combines time series data with other unstructured data. The embodiment of the present disclosure can break through the limitations of a single data source, combine comprehensive analysis of data from different fields, and improve the accuracy of time series prediction.

[0020] The disclosed embodiments determine a third selection indicator associated with a first selection indicator based on at least one public-domain data information indicator, and determine individuals in a second time series dataset based on the first, second, and third selection indicators. This allows for in-depth analysis of potential relationships between different data sources and the discovery of connectivity between indicators across these data sources, thereby further improving the prediction accuracy of time series data. BRIEF DESCRIPTION OF THE DRAWINGS

[0021] The above contents and other objects, features and advantages of the present disclosure will become more apparent through the following description of the embodiments of the present disclosure with reference to the accompanying drawings, in which:

[0022] Figure 1 The application scenario diagram of the domain-based indicator selection strategy construction system, method, device, medium and program product according to the embodiment of the present disclosure is schematically shown.

[0023] Figure 2 A block diagram of a system for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0024] Figure 3 The following schematically shows a block diagram of a system for constructing a second domain-based indicator selection strategy according to an embodiment of the present disclosure.

[0025] Figure 4 A block diagram of a system for constructing a third domain-based indicator selection strategy according to an embodiment of the present disclosure is schematically shown.

[0026] Figure 5 A flowchart of a method for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0027] Figure 6 The flowchart of the second method for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0028] Figure 7 The flowchart of the third method for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0029] Figure 8The flowchart of the fourth method for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0030] Figure 9 The figure shows the backtest return performance of the entire domain stocks as holding stocks from 2014 to 2024 according to the embodiment of the present disclosure.

[0031] Figure 10 The final backtest yield curve according to an embodiment of the present disclosure is shown.

[0032] Figure 11 A block diagram of an electronic device suitable for implementing a method for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0033] It should be noted that, for the sake of clarity, in the drawings used to describe the embodiments of the present disclosure, the sizes of the overall / local structures or overall / local areas may be enlarged or reduced, that is, these drawings are not drawn according to the actual scale. DETAILED DESCRIPTION

[0034] Hereinafter, embodiments of the present disclosure will be described with reference to the accompanying drawings. However, it should be understood that these descriptions are merely exemplary and are not intended to limit the scope of the present disclosure. In the detailed description below, for ease of explanation, many specific details are set forth to provide a comprehensive understanding of the embodiments of the present disclosure. However, it is apparent that one or more embodiments may also be implemented without these specific details. In addition, in the following description, descriptions of well-known structures and technologies are omitted to avoid unnecessary confusion of the concepts of the present disclosure.

[0035] The terms used herein are only for describing specific embodiments and are not intended to limit the present disclosure. The terms "comprise," "include," etc. used herein indicate the presence of the features, steps, operations, and / or components, but do not exclude the presence or addition of one or more other features, steps, operations, or components.

[0036] All terms used herein (including technical and scientific terms) have the meanings commonly understood by those skilled in the art unless otherwise defined. It should be noted that the terms used herein should be interpreted as having a meaning consistent with the context of this specification and should not be interpreted in an idealized or overly rigid manner.

[0037] When expressions such as "at least one of A, B, and C, etc." are used, they should generally be interpreted in accordance with the meaning commonly understood by those skilled in the art (for example, "a system having at least one of A, B, and C" should include but is not limited to a system having A alone, B alone, C alone, A and B, A and C, B and C, and / or A, B, C, etc.).

[0038] In the actual processing of time series data, even seemingly simple user strategy requirements can trigger extensive computation. For example, a user might want to retrieve a certain type of data. If the data volume is enormous, potentially involving tens of thousands or even more data points, this requires the computer to perform extensive numerical calculations and logical analysis. Comprehensive screening of multiple indicators, comparative analysis of data from different time periods, and other such processes will undoubtedly further increase the complexity and workload of the calculations, placing even higher demands on the computer's computing power and processing efficiency. Ordinary users often lack the ability to discern the potential connections between different strategy indicators. Therefore, in real-world applications, the user's screening strategy often fails to produce the expected results.

[0039] Time series analysis (such as trend analysis, cyclical fluctuations, and outlier detection) can uncover patterns in time series data to aid decision-making. For example, predicting long-term warming trends or extreme weather cycles in climate change; predicting the degradation rate and service life of material properties; and signaling the continuation of stock price trends (such as momentum effects) or reversals. Causal inference methods from climate science can be used to analyze the validity of stock market indicators; and time series models used to process wear data in materials science can be transferred to predict stock drawdown risk. Understanding these commonalities helps break down domain barriers and enable a more universal approach to analyzing data patterns in complex systems.

[0040] In the process of implementing the present invention, the inventors discovered that the existing technology has at least the following problems: ordinary time series data users lack scientific and effective tools and methods to analyze and make decisions on massive time series data. The amount of data required for comparative analysis of the entire data is large, making it difficult to perform multi-parameter indicator analysis and recommendations; data from different fields cannot be comprehensively analyzed to explore the potential patterns between changes in data in different fields. The related technology only achieves segmented investment by simply filtering out stocks with poor quality or extreme prices. It lacks scientific and effective methods to fully utilize the segmented stock data, resulting in low efficiency in mining segmented stocks; it fails to effectively utilize the characteristics of stock time series data to explore and utilize the patterns of stock changes over time. Secondly, the existing technology has not yet disclosed a system that can effectively backtest strategies based on indicator segmentation.

[0041] Time series analysis (such as trend analysis, cyclical fluctuations, and outlier detection) can uncover patterns in time series data to aid decision-making. For example, predicting long-term warming trends or extreme weather cycles in climate change; predicting the degradation rate and service life of material properties; and signaling the continuation of stock price trends (such as momentum effects) or reversals. Causal inference methods from climate science can be used to analyze the validity of stock market indicators; and time series models used to process wear data in materials science can be transferred to predict stock drawdown risk. Understanding these commonalities helps break down domain barriers and enable a more universal approach to analyzing data patterns in complex systems.

[0042] The embodiment of the present disclosure provides a domain-based indicator selection strategy construction system, which is characterized by comprising:

[0043] The data import module is configured to, in response to obtaining a first selection indicator from a client, call a first data interface based on the first selection indicator to determine and import a second time series data set within a first time interval from the first time series data set.

[0044] The indicator calculation module is configured to perform indicator calculation on the second time series data set based on the first selected indicator and a second indicator set divided into different categories from the first selected indicator to obtain a first indicator calculation result.

[0045] The indicator analysis module is configured to perform indicator analysis calculation according to the first indicator calculation result to obtain the first indicator analysis result.

[0046] The domain analysis module is configured to determine a second selection indicator associated with the first selection indicator from a second indicator set based on the first indicator analysis result and the second time series data set.

[0047] Figure 1 The following schematically illustrates an application scenario diagram of a domain-based indicator selection strategy construction system, method, device, medium, and program product according to an embodiment of the present disclosure. It should be noted that: Figure 1 What is shown are merely examples to which the embodiments of the present disclosure may be applied, to help those skilled in the art understand the technical content of the present disclosure, but does not mean that the embodiments of the present disclosure cannot be used in other devices, systems, environments or scenarios.

[0048] like Figure 1 As shown, the application scenario 100 according to this embodiment may include a first terminal device 101, a second terminal device 102, a third terminal device 103, a network 104, and a server 105. The network 104 is used as a medium for providing a communication link between the first terminal device 101, the second terminal device 102, the third terminal device 103, and the server 105. The network 104 may include various connection types, such as wired or wireless communication links or optical fiber cables.

[0049] Users can use a first terminal device 101, a second terminal device 102, or a third terminal device 103 to interact with a server 105 via a network 104 to receive or send messages, etc. Various communication client applications can be installed on the first terminal device 101, the second terminal device 102, and the third terminal device 103, such as shopping applications, web browser applications, search applications, instant messaging tools, email clients, social platform software, etc. (for example only). The first terminal device 101, the second terminal device 102, and the third terminal device 103 can also be referred to as clients.

[0050] The first terminal device 101 , the second terminal device 102 , and the third terminal device 103 may be various electronic devices having display screens and supporting web browsing, including but not limited to smart phones, tablet computers, laptop computers, desktop computers, and the like.

[0051] Server 105 can be a server that provides various services, such as a backend management server (for example only) that supports websites browsed by users using first terminal device 101, second terminal device 102, and third terminal device 103. The backend management server can analyze and process received user requests and other data, and provide feedback (e.g., web pages, information, or data obtained or generated based on user requests) to the terminal devices. For example, server 105 can be a standalone physical server, a server cluster or distributed system composed of multiple physical servers, or a cloud server that provides basic cloud computing services such as cloud services, cloud computing, network services, and middleware services.

[0052] In applications such as client applications and web applications (APPs), the client (i.e., front-end) and the server (i.e., back-end) can communicate data through network messages. For example, in the APP client, the server assembles the parameters that need to be obtained from the client and calls the network request method to send them to the server.

[0053] It should be noted that the domain-based indicator selection strategy construction method provided in the embodiment of the present disclosure can generally be executed by at least one of the terminal device or the server. Accordingly, the domain-based time series data set selection device provided in the embodiment of the present disclosure can generally be set in at least one of the terminal device or the server. The domain-based time series data set selection method provided in the embodiment of the present invention can also be executed by a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105. Accordingly, the domain-based time series data set selection device provided in the embodiment of the present invention can also be set in a server or server cluster that is different from the server 105 and can communicate with the first terminal device 101, the second terminal device 102, the third terminal device 103 and / or the server 105.

[0054] It should be understood that Figure 1 The number of terminal devices, networks and servers in the embodiment is merely illustrative. Any number of terminal devices, networks and servers may be provided as required.

[0055] The following will be based on Figure 1 The described scenario provides a detailed description of a domain-based indicator selection strategy construction system, method, device, medium, and program product according to an embodiment of the present disclosure.

[0056] Figure 2 A block diagram of a system for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0057] like Figure 2 As shown, this embodiment includes:

[0058] The data import module 220 is configured to, in response to obtaining a first selection index from the client, call the first data interface 210 based on the first selection index to determine and import a second time series data set within a first time interval from the first time series data set.

[0059] In meteorological research, to predict changes in tropical rainforests, the world or a specific region can be divided into different climate zones, such as tropical rainforest climate zones and temperate continental climate zones, based on climate factors such as temperature, precipitation, and air pressure. For example, if the first selected indicator is the tropical rainforest indicator, meteorological data related to tropical rainforests can be filtered out from global meteorological data.

[0060] In the field of stock finance, stock data can be divided into small-cap stocks and other types according to the stock market situation. In an embodiment of the present disclosure, a first selection indicator input from a target client can be obtained in response to a financial platform. The first selection indicator often represents the target client operator's construction of a stock selection strategy or reflects his or her investment preferences, etc. The strategy may also include: the time to buy stocks (opening or closing), the method of adjusting positions, the number of stocks selected, etc. Frequency of position changes: can be set to weekly, monthly, etc. Number of stocks held: a fixed number is required for each period, such as 5, 10, 30, etc. Ordinary operators often have a superficial understanding of stock selection strategy indicators, and it is difficult to obtain in-depth indicators that have similar effects to the selected strategy indicators.

[0061] The first selected indicator can be, for example, market capitalization, liquidity, turnover rate, reversal, momentum and other indicators. These indicators are various variables or characteristics that affect stock returns. There may be correlation, complementarity, or overlap between different indicators. Based on the first selected indicator, stocks in the sub-domain related to the first selected indicator can be selected from the entire stock pool as the second time series data set, such as using the small market capitalization indicator to select small market capitalization stocks. The stock pool data source can be locally stored or the stock data of the corresponding time period can be obtained from the database. The size of the sub-domain stock pool should generally be at least larger than the number of stocks in the target holding stock pool.

[0062] In this embodiment, in response to the target client inputting a request to build a strategy into the system, the system calls a data interface based on the interface, selects a strategy based on the user's time series data, and extracts stock volume and price data, financial data, or macroeconomic fundamentals data within the target time period from the associated database. The disclosed embodiment can be combined with a variety of financial data interfaces, the free and open Akshare interface, or the commercial Wind data API interface. The data can be processed before use, including: standardization of stock data, including standardization of date formats. Forward or backward filling is used to handle missing data.

[0063] The first time interval can be a period of time including a start time and an end time, serving as the backtesting period. This first time interval can be pre-configured or input by the target client into the financial platform. Based on the first time interval, the system calls a data interface to extract stock volume and price data, financial data, or macroeconomic fundamentals data for the target time period from the associated database. The first time interval, also known as the backtesting period, can be flexibly set to one month, one year, or ten years, based on local data or available time periods in the data interface.

[0064] In this embodiment, in response to obtaining a first selection indicator from the client, the data import module 220 calls the first data interface based on the first selection indicator to determine and import the second time series data set from the first time series data set within the first time interval, thereby avoiding importing the entire data for calculation and reducing the amount of data calculation.

[0065] The indicator calculation module 230 is configured to perform indicator calculation on the second time series data set based on the first selected indicator and a second indicator set classified into a different category from the first selected indicator to obtain a first indicator calculation result;

[0066] The first selection indicator can usually be used to perform domain operations on time series data, can be used to filter data, and can further reduce the scale and computational complexity of data processing.

[0067] In meteorological research, a region is classified as either a "rainy" or "dry" climate based on precipitation observations. A second set of indicators, such as rainfall, land and sea location, and topography, which are categorized differently from the climate type indicators, is used to perform indicator analysis and calculations on a second time series dataset within a first time interval to obtain the first indicator analysis and calculation results.

[0068] In the stock market, in embodiments of the present disclosure, if the first selected indicator is a small-cap indicator based on market capitalization, other indicators outside of the market capitalization category can be used as the second indicator set. For example, the second indicator set can be style-based indicators. Such indicators may include value, momentum, volatility, growth, earnings, liquidity, leverage, and reversal indicators.

[0069] The indicator calculation module 230 may be configured to calculate time series indicators (usually based on data of a single individual) and / or cross-sectional indicators (usually based on data of multiple individuals).

[0070] For example, in the meteorological research field, the indicator analysis and calculation module can calculate indicators such as rainfall changes in a region and the distribution of rainfall in different regions at the same time. In the stock market indicator analysis and calculation module, you can first calculate time-series indicators (usually based on individual stock data), then use data from the entire stock pool to calculate cross-sectional indicators, and then further analyze the calculated indicators. The order of calculation can also be adjusted.

[0071] The indicator analysis module 240 is configured to perform an indicator analysis calculation based on the first indicator calculation result to obtain a first indicator analysis result;

[0072] The results calculated by the indicator calculation module are imported into the indicator analysis module. For example, one or more of the following analyses can be performed: a) RankIC vector calculation; b) group net value vector calculation; c) indicator style exposure vector calculation.

[0073] The domain analysis module 250 is configured to determine a second selection indicator associated with the first selection indicator from the second indicator set based on the first indicator analysis result and the second time series data set. In the embodiment of the present disclosure, the second selection indicator can be one or more.

[0074] In the field of meteorological research, the domain analysis module combines the first selection index and the second selection index, and in other embodiments of the present disclosure, may further combine the third selection index and the fourth selection index to perform the following analysis:

[0075] a) Comprehensive change curve of meteorological indicators within the region: Analyze the time-varying curve of core meteorological indicators such as temperature, air pressure, and humidity after comprehensive weighting within a specific meteorological region to explore the evolution trend of the meteorological environment in the region. For example, the changes in the comprehensive curve of monthly average climate indicators in a monsoon climate region can be used to determine abnormal climate fluctuations.

[0076] b) Changes in the number of meteorological stations and the distribution of monitoring data over time within the domain: Statistics on the increase and decrease in the number of monitoring stations in each meteorological sub-domain, as well as the distribution characteristics of different types of meteorological data (such as precipitation, wind speed, etc.) in the temporal dimension. For example, the changes in the number of precipitation monitoring stations in arid areas and the differences in the distribution of precipitation data at each station at different time periods can be studied.

[0077] c) Information coefficient (IC) performance of meteorological element correlation within the domain: Calculate the information coefficient between different meteorological elements within the domain (such as temperature and evaporation, wind speed and dust concentration, etc.), evaluate the strength and stability of linear correlation between elements, and identify element associations that play a key role in regional meteorological changes.

[0078] d) Average quantile performance of the correlation between meteorological elements within and outside the region: The average quantile of the correlation between meteorological elements within and outside the region is calculated respectively, and the differences in the degree of correlation between meteorological elements in different regions are compared and analyzed.

[0079] In the stock market, based on a specific domain strategy, the domain module is called to combine the first and second selection indicators. In other embodiments of the present disclosure, the third and fourth selection indicators may be further combined to perform the following analysis on the domain stock pool:

[0080] a) Curve performance of funds held within a sector. For example, within a specific stock sector, we can analyze the weighted changes in core indicators such as stock price, price-to-earnings ratio, and price-to-book ratio over time to explore the evolving trends of the stock market within that sector. For example, for stocks in the new energy vehicle sector, we can analyze the weighted changes in the quarterly average stock price, price-to-earnings ratio, and other indicators to identify abnormal fluctuations in that sector.

[0081] b) Changes in the number of stocks and market capitalization within a domain over time. For example, we can analyze the increase or decrease in the number of constituent stocks within each stock sub-domain, as well as the temporal distribution characteristics of stocks of different market capitalizations. For example, within the small-cap sub-domain, we can analyze the changes in the number of stocks with a market capitalization below 10 billion yuan, as well as the differences in the market capitalization distribution of these stocks across different quarters.

[0082] c) IC performance of style indicators within a domain. For example, calculate the information coefficient between different stock indicators within the domain (such as growth indicators, value indicators, momentum indicators, etc.), evaluate the strength and stability of linear correlations between indicators, and identify indicator correlations that are key to stock price fluctuations within the subdomain.

[0083] d) Average quantile performance of style indicators within and outside the domain. For example, calculate the average quantile of the correlation between stock indicators within and outside the domain for each stock, and compare and analyze the differences in the degree of correlation between stock indicators in different domains.

[0084] This embodiment determines a second selection indicator associated with the first selection indicator from the second indicator set based on the first indicator analysis result and the second time series data set. This embodiment further improves data screening processing efficiency by obtaining the first selection indicator and the second selection indicator.

[0085] Figure 3 The following schematically shows a block diagram of a system for constructing a second domain-based indicator selection strategy according to an embodiment of the present disclosure.

[0086] like Figure 3 As shown, the system of this embodiment further includes:

[0087] The second data interface 260 is configured to crawl public data related to individuals in the second time series dataset from public data within a first time interval according to a call. In one embodiment, the public data can be internet data or other public data sources different from the first time series dataset. The second data interface is called to crawl public data related to individuals in the second time series dataset from public data.

[0088] In the field of meteorological research, you can crawl from the Internet to obtain information related to the El Niño / La Niña phenomenon, human activities, and other individual meteorological time series data.

[0089] In the stock market, this can be done by crawling macroeconomic information, industry information, company information, media and public opinion information, etc. related to individual time series data from Internet data; based on macroeconomic information, industry information, company information, media and public opinion information, etc., determine the degree of influence of public data information on the individuals in the second time series data set.

[0090] The system in this embodiment further includes: an information analysis module 270, configured to determine a public data analysis result of an individual of the second time series dataset based on the public data information.

[0091] The indicator analysis module 240 is further configured to perform indicator analysis calculation based on the public data analysis result to obtain a second indicator analysis result.

[0092] Because potential connections exist between indicator parameters in data from different domains, calculating public data information indicators within public data can exploit these connections. For example, the "herd effect" of internet information dissemination can amplify short-term sentiment, causing stock prices to deviate from fundamentals, thus providing a data foundation for reversal indicators. Examples of public data information indicators within public data include business development indicators, management change indicators, and financial disclosure indicators. For example, analysis reveals that when the first-selected indicator is the "small market capitalization" indicator, the business development indicator has a more significant impact on it than the management change indicator and the financial disclosure indicator. Therefore, the "business development" indicator can be considered a third-selected indicator associated with the "small market capitalization" indicator. By using multiple potentially correlated indicators such as "small market capitalization," "reversal," and "business development," the efficiency of stock time series screening can be improved. In this embodiment, at least one public data information indicator for an individual in the second time series dataset can be determined based on public data information using an indicator analysis and calculation approach. For example, the RankIC vector value of the "business development" indicator can be calculated.

[0093] In the field of meteorological research, for example, human activities may produce greenhouse gas emissions and aerosol emissions, causing large-scale deforestation and marine pollution, thereby having potential impacts on the local climate.

[0094] In the stock market, for example, from company financial news, metrics such as revenue growth rate and net profit margin can be extracted; from social media discussions, the frequency of specific keywords such as "buy" and "sell" can be counted. Alternatively, public opinion terms can be selected from online search requests to create a "stock public opinion index," and a buy signal can be issued when the public opinion index meets a certain pattern.

[0095] The domain analysis module 250 is further configured to determine a third selection indicator associated with the first selection indicator based on the second indicator analysis result.

[0096] In meteorological research, based on the results of the second indicator analysis and calculation, rainfall indicators closely related to tropical rainforests can be selected as the third selected indicator. For example, the results of tropical rainforest, rainfall labeling, and human activity analysis and calculation can be used as selected indicators after the optimization strategy, and individuals in the second time series dataset can be selected as the subjects of meteorological change research.

[0097] In this example of the stock market, based on the analysis and calculation results of the second indicator, the indicator with the best stock performance in the second indicator set is selected as the third selection indicator. The first selection indicator, the second selection indicator, and the third indicator are analyzed and calculated to form the selection indicator after the optimization strategy is implemented. Individuals in the second time series dataset are determined and used as recommended stocks. In the embodiments of the present disclosure, the third selection indicator can be one or more.

[0098] In some embodiments of the present disclosure, the indicator calculation module is further configured to perform indicator analysis and calculation on the first time series data set based on a fourth selection indicator set associated with the second selection indicator to obtain a third indicator analysis and calculation result; the domain analysis module is further configured to determine a fourth selection indicator associated with the first selection indicator from the fourth selection indicator set based on the third indicator analysis and calculation result.

[0099] In this embodiment, a certain type of indicator can further include related sub-indicators. For example, in the stock market, reversal indicators can be classified according to different categories, including reversal indicators classified by time period, reversal indicators classified by market-driven logic, etc. Different indicators perform differently in a certain sub-domain. Through this embodiment, it is possible to further obtain indicators related to the target strategy, further optimize the stock selection strategy, and improve stock selection efficiency.

[0100] Based on the results of the third indicator analysis, indicators that meet the absolute value of the RankIC mean and / or have a high IC win rate can be selected as the fourth selection indicator. Based on the infinite norm of the indicator style exposure vector, the indicators that meet the conditions are divided into the corresponding second indicator set. The individuals of the second time series data set are determined based on the first selection indicator, the second selection indicator, and the fourth selection indicator. In the embodiments of the present disclosure, the fourth selection indicator can be one or more.

[0101] In meteorological research, by calculating the average quantile vectors of the correlation between meteorological elements within and outside the domain, as well as the infinite norm of the difference between the two, corresponding meteorological element categories are extracted as key influencing indicators in meteorological forecasting and climate research strategy construction, providing a core basis for meteorological disaster warnings and climate model optimization. In the stock market, by calculating the average quantile vectors within and outside the domain, as well as the infinite norm of the difference between the two, corresponding style indicator categories are extracted as key indicators in strategy construction.

[0102] Figure 4 A block diagram of a system for constructing a third domain-based indicator selection strategy according to an embodiment of the present disclosure is schematically shown.

[0103] like Figure 4As shown, the system also includes a strategy construction module 280, which is configured to determine the individuals of the second time series data set based on the first selection indicator and the second selection indicator. In the strategy construction module, the domain analysis module can be called to obtain the report data set under the indicator domain; according to the stock domain conditions, the target stock pool is screened out from the entire stock pool. In this embodiment, in the meteorological research field, in the strategy construction module, the target meteorological data observation pool can be screened out from the global meteorological observation data pool. In the stock field, in the strategy construction module, according to the stock domain conditions, the target stock pool can be screened out from the entire stock pool.

[0104] In some embodiments of the present disclosure, the system may further include a parameter setting module, in which strategy parameters including the following may be configured: a) strategy path (strategy to be called); b) start time and end time of backtesting; c) time to buy stocks (opening or closing); d) position adjustment method; e) number of stocks to be selected.

[0105] In some embodiments of the present disclosure, the system may further include a backtesting module. The second selection indicator, the third selection indicator, and the fourth selection indicator in the above embodiments are used as target style indicators, and all target style indicators are imported into the backtesting module one by one to perform historical return backtesting. Based on the sorting results, the indicator with the best return is selected as the target indicator for strategy construction. According to the target indicator, its positive / negative prediction effect is determined in the indicator analysis module. On each position adjustment date, the indicator value of each stock is sorted in descending / ascending order, and the final stock holdings are determined in combination with the stock selection quantity parameter. The selected high-quality style holdings are pushed to target customers through the financial platform to provide them with accurate investment decision support.

[0106] For the parts not mentioned in the apparatus part, they can be understood with reference to the various embodiments of the above-mentioned method. That is, the apparatus part includes modules for executing the various steps of any one of the method embodiments described above. In addition, the implementation methods, technical problems solved, functions achieved, and technical effects achieved of each module / unit / subunit, etc. in the apparatus part embodiment are respectively the same or similar to the implementation methods, technical problems solved, functions achieved, and technical effects achieved of each corresponding step in the method part embodiment, and will not be repeated here.

[0107] According to an embodiment of the present disclosure, any multiple modules among the data import module 220, the indicator calculation module 230, the indicator analysis module 240, the domain analysis module 250, the information analysis module 270, and the strategy construction module 280 can be combined into a single module for implementation, or any one of these modules can be split into multiple modules. Alternatively, at least part of the functionality of one or more of these modules can be combined with at least part of the functionality of other modules and implemented in a single module.

[0108] According to an embodiment of the present disclosure, at least one of the data import module 220, the indicator calculation module 230, the indicator analysis module 240, the domain analysis module 250, the information analysis module 270, and the strategy construction module 280 can be at least partially implemented as a hardware circuit, such as a field programmable gate array (FPGA), a programmable logic array (PLA), a system on a chip, a system on a substrate, a system on a package, an application-specific integrated circuit (ASIC), or can be implemented by hardware or firmware in any other reasonable way of integrating or packaging the circuit, or implemented in any one of the three implementation methods of software, hardware, and firmware, or in an appropriate combination of any of them. Alternatively, at least one of the data import module 220, the indicator calculation module 230, the indicator analysis module 240, the domain analysis module 250, the information analysis module 270, and the strategy construction module 280 can be at least partially implemented as a computer program module, which can perform the corresponding function when the computer program module is executed.

[0109] Based on the above domain-based indicator selection strategy construction system, the present disclosure also provides a domain-based indicator selection strategy construction method. The device will be described in detail below with reference to the accompanying drawings.

[0110] Figure 5 A flowchart of a method for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0111] like Figure 5 As shown, this embodiment includes:

[0112] In operation S510, in response to obtaining a first selection index from a client, the data import module calls a first data interface based on the first selection index to determine and import a second time series data set within a first time interval from the first time series data set;

[0113] In operation S520, the indicator calculation module performs a first indicator calculation on the second time series data set based on the first selected indicator and a second indicator set classified into a different category from the first selected indicator to obtain a first indicator calculation result;

[0114] In operation S530, the first indicator calculation result is imported into the indicator analysis module to perform indicator analysis calculation to obtain the first indicator analysis result;

[0115] In some embodiments of the present disclosure, the indicator analysis calculation includes at least one of: information coefficient calculation, ranking information coefficient calculation, and style indicator exposure analysis. The second indicator set is , call the indicator analysis module.

[0116] (1) Information coefficient (IC) calculation.

[0117] Assume that the time series of the backtest interval is , the indicator value sequence is , the return series of the benchmark index , calculate IC:

[0118] (1.1)

[0119] in It is the difference between the index value and the ranking of future benchmark returns.

[0120] (2) Calculation of ranking information coefficient (RankIC vector).

[0121] RankIC Vector ,in:

[0122] (1.2)

[0123] (1.3)

[0124] (1.4)

[0125] (1.5)

[0126] The IC mean ( ) measures the overall predictive power of the indicator over the backtest period; higher values are better. ICWR (ICWin Ratio) indicates the proportion of periods with a positive IC; higher values indicate more stable indicator predictions.

[0127] (3) Style indicator exposure analysis.

[0128] Style indicator exposure analysis is mainly used to analyze the tendency of a certain indicator to belong to the indicator type of the second indicator set. Calculate the indicator style exposure vector ,in:

[0129] (1.6)

[0130] Here, the infinite norm of the style exposure vector, i.e., the corresponding max|S|, represents the style attribute i of the indicator. Common style indicators, according to Barra's style indicators, include market capitalization, value, momentum, volatility, growth, earnings, liquidity, leverage, and reversal.

[0131] In operation S540 , the domain analysis module determines a second selection indicator associated with the first selection indicator from the second indicator set based on the first indicator analysis result and the second time series data set.

[0132] In some embodiments of the present disclosure, determining a second selection indicator associated with the first selection indicator from a second indicator set based on the first indicator analysis and calculation result may be implemented in the following manner.

[0133] For the stock pool after the stocks are segmented according to the first selection index, the segmentation analysis module performs segmentation analysis. Assume that the time series of the number of stocks in the stock pool is , at least guarantee , that is, the number of stocks in the sub-domain stock pool must always be greater than the number of stocks held. The RankIC matrix of intra-domain style indicators is calculated based on the second indicator set, and each style indicator is an alternative second-selection indicator.

[0134] (1.7)

[0135] in, is the IC value of the i-th style indicator in the domain; The IC value of the i-th style indicator in the entire market; The absolute value change of IC of the i-th style indicator. The style indicator corresponding to the maximum value is used as the strategy indicator.

[0136] Calculate the strategy indicators for the stocks in the domain and get the strategy indicator matrix:

[0137] (1.8)

[0138] At each rebalancing point t, the indicator analysis module is called to analyze the RankIC vector symbols of the strategy indicators, sort them, and select the number of stocks requested by the customer. Assume that the number of stocks in the domain is M, and the number of stocks held is fixed at N.

[0139] if , that is, the sign of IC is negative, sort in ascending order of indicators, and select the indicator vector at time t The smallest N stocks in the list are used as holding stocks; on the contrary, if , the sign of IC is positive, and the indicators are arranged in descending order. The N stocks with the largest index vector are selected as holdings.

[0140] In addition, if the IC of a certain style indicator q is positive / negative, the stocks in the domain stock pool are arranged in descending / ascending order, and the top N stocks are selected as recommended holdings.

[0141] In this example, the following analysis tools are further proposed to enhance the understanding of indicator performance:

[0142] 1) RanKIC Monthly Heat Map: This chart visually displays the performance of the indicator in different months, helping investors understand the timeliness and volatility of the indicator.

[0143] 2) Grouped net worth chart: Analyze the monotonicity of the indicator to measure its interpretability.

[0144] 3) Grouped holding trends: Observe the performance of returns in different holding periods to facilitate the management of holding periods.

[0145] 4) Indicator Style Exposure Analysis: Identify and avoid indicator crowding risks by calculating the correlation between the indicator and Barra Common Style Indicators.

[0146] 5) Industry RankIC chart: evaluates the indicator's ability to predict earnings in various industries, providing effective support for industry rotation strategies.

[0147] 6) Market Capitalization Share and Market Capitalization RankIC: Analyze the market capitalization style characteristics of the indicator to optimize the performance of the target indicator by combining the market capitalization indicator domain.

[0148] Figure 6 The flowchart of the second method for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0149] like Figure 6 As shown, after executing operation S540, this embodiment includes:

[0150] In operation S610, a second data interface is called to crawl public data information related to an individual in a second time series dataset within a first time interval from public data;

[0151] In operation S620, the information analysis module determines a public data analysis result of the second time series dataset individual based on the public data information;

[0152] In operation S630, the public data analysis result is imported into the index analysis module to perform index analysis calculation to obtain a second index analysis result;

[0153] In operation S640 , the domain analysis module determines a third selection index associated with the first selection index based on the second index analysis result.

[0154] Figure 7 The flowchart of the third method for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0155] like Figure 7 As shown, after executing operation S540, this embodiment includes:

[0156] In operation S550 , the strategy construction module determines individuals of the second time series data set based on the first selection index and the second selection index.

[0157] In this embodiment, to verify the effectiveness of the optimized strategy in the regional stock pool, the return and risk of the strategy need to be evaluated. The return and risk of the strategy can be evaluated based on the return calculated by the first selection indicator and the second selection indicator in the second time series data set.

[0158] In this embodiment, the following core indicators are used:

[0159] Assume that the initial capital is , the investment period is , the rate of return in period t is .

[0160] (1) Cumulative net value:

[0161] (1.9)

[0162] (2) Annualized rate of return:

[0163] (1.10)

[0164] (3) Maximum drawdown:

[0165] Maximum drawdown measures the maximum loss of a portfolio. The specific steps are as follows:

[0166] Calculate the cumulative net value series:

[0167] (1.11)

[0168] Calculate the historical maximum net value:

[0169] (1.12)

[0170] Calculate the retracement size at the current time point t:

[0171] (1.13)

[0172] Maximum drawdown indicator:

[0173] (1.14)

[0174] This indicator represents the maximum drop in net value at a certain moment compared to the historical highest net value. The smaller the drawdown, the lower the risk of the strategy and the more stable the return.

[0175] (4) Winning rate

[0176] The winning rate indicates the proportion of profitable periods in the total investment period:

[0177] (1.15)

[0178] (5) Annualized Return / Drawdown Ratio (Calmar Ratio)

[0179] The Calmar ratio measures the risk-adjusted return of a strategy and is defined as:

[0180] (1.16)

[0181] A higher Calmar ratio indicates a strategy's ability to achieve higher returns with lower risk.

[0182] In this embodiment, the second selection indicator, the third selection indicator, the fourth selection indicator and other relevant selection indicators can be used as the target style indicator library, and all the indicators therein can be imported one by one for historical return backtesting. Based on the sorting results, the indicator with the best return is selected as the target indicator for strategy construction. According to the target indicator, its positive / negative prediction effect is determined in the indicator analysis module. On each position adjustment date, the indicator value of each stock is sorted in descending / ascending order, and the final stock holdings are determined in combination with the stock selection quantity parameter. The selected high-quality style holdings are pushed to the target customers through the financial platform to provide them with accurate investment decision-making support. By importing the indicators one by one for historical return backtesting, the accuracy of indicator selection can be further improved, and the efficiency of strategy selection construction can be improved.

[0183] In some embodiments of the present disclosure, a series of analysis tools for stocks within a domain may be provided, specifically including:

[0184] 1) Select all fund curves in the domain: Track the fund curve changes of all stocks in the domain.

[0185] 2) Time series distribution of stock quantity: helps determine the number of stocks to be selected for each portfolio adjustment.

[0186] 3) Time series distribution of stock pool market capitalization: Analyze the overall market capitalization percentiles of stocks in the domain and the market capitalization proportions of large and small companies.

[0187] 4) Stock pool industry distribution analysis: Examine the industry distribution characteristics of stocks in the domain in each year.

[0188] 5) Analysis of the performance of style indicators within the domain: Identify style indicators with outstanding performance within the domain to help build more effective quantitative strategies.

[0189] 6) Comparison of the performance of the intra-domain and full-market style indicators IC: By comparing the performance within the domain and the entire market, identify indicators with significant excess returns within the domain to improve the effectiveness of the indicator domain segmentation strategy.

[0190] 7) Percentile performance of style indicators outside the domain: The larger the gap between the percentile of the indicator within the domain and outside the domain, the stronger the excess return potential of the indicator within the domain, which is equivalent to amplifying the return performance of the indicator, thereby improving the investment return of the strategy.

[0191] Through the above solution, this embodiment significantly improves the stock selection efficiency of quantitative investment strategies in a specific stock pool and the scientific nature of strategy backtesting, providing investors with a more accurate decision support tool.

[0192] In some embodiments of the present disclosure, the holding stock screening method may include: (1) determining the excess return level of each stock based on the numerical value of the strategy indicator. (2) judging the prediction direction of the indicator based on the RankIC sign of the strategy indicator: a. Positive prediction indicator: sort in descending order and select the top N stocks as the holding stock pool. b. Negative prediction indicator: sort in ascending order and select the bottom N stocks as the holding stock pool.

[0193] The disclosed embodiments optimize target stock portfolios through domain-specific analysis of indicators, combining the characteristics of style and domain-specific indicators. This method significantly improves the utilization efficiency of stock data and enhances the accuracy and intelligence of financial platform recommendation results.

[0194] It should be noted that some steps of the above method can be executed individually or in combination, and can be executed in parallel or sequentially, and are not limited to the specific operation sequence shown in the figure.

[0195] Figure 8 The flowchart of the fourth method for constructing an indicator selection strategy based on domain division according to an embodiment of the present disclosure is schematically shown.

[0196] like Figure 8 As shown, in the parameter configuration module, the client can set parameters such as call strategy, backtesting time range, position adjustment method, holding period, number of stocks to be selected, etc.

[0197] In response to receiving the backtesting time and strategy data requirements input by the target client to the financial platform, the first data interface is called.

[0198] Read the name of the first-selected indicator used in the strategy file, match it to the corresponding indicator file, and call the indicator calculation module. In the indicator calculation module, calculate the required indicators one by one. You can first calculate the time series indicator (usually based on individual stock data), and then use the data of the entire stock pool to calculate the cross-sectional indicator.

[0199] Call the indicator analysis module, select target indicators that meet customer needs based on the results of the first indicator analysis calculation, build an indicator library, and import the calculated indicators into the indicator analysis module. Perform RankIC vector calculations, grouped net value vector calculations, and indicator style exposure vector calculations. Based on the output parameters, select target indicators that meet customer needs and build an indicator library.

[0200] A stock-segment strategy is constructed using the indicator library. The segmentation analysis module is then invoked to analyze the characteristics of the segmented stock pool, selecting suitable indicators. The module analyzes the following aspects of the segmented stock pool: a) the performance of the fund curve within the segment; b) the change in the number of stocks and market capitalization distribution over time within the segment; c) the IC performance of style indicators within the segment; and d) the average quantile performance of style indicators within and outside the segment. The module then calculates the average quantile vectors within and outside the segment, as well as the infinite norm of their difference, and extracts the corresponding style indicator categories as key indicators for strategy construction.

[0201] The indicators in the backup indicator library are linearly combined to construct a strategy. The backtesting system is called to filter out the second-choice indicators required by the customer and determine the strategy plan.

[0202] This example demonstrates the modular relationship between parameter input, data acquisition, factor analysis, domain analysis, and backtesting. This process fully utilizes the molecular domain strategy to improve the efficiency of stock data mining and the stability of stock selection strategies.

[0203] Since stock data is easier to quantify, it will be further described below with reference to a specific embodiment of the present disclosure in the stock field.

[0204] To facilitate a better understanding of the concept of the embodiments of the present disclosure and one or more specific implementation details, the terms involved in some embodiments of the present disclosure are first explained as follows:

[0205] RankIC (Rank Information Coefficient): A statistical measure of a stock selection indicator's ability to rank an asset's future returns. RankIC, based on the Spearman rank correlation coefficient, reflects the correlation between the indicator's predicted ranking and its actual return ranking. Higher values indicate a better indicator's ability to predict excess returns.

[0206] Barra Style Factors: Developed by Barra, these are core variables in a multi-metric model widely used in risk modeling and performance attribution analysis in quantitative investing. Barra Style Factors reveal the sources of return and risk characteristics of a stock or portfolio by decomposing systematic risk indicators.

[0207] Market capitalization: A measure of company size, usually expressed as logarithmic market capitalization. Small-cap stocks tend to have higher risk premiums and exhibit stronger growth potential.

[0208] Value metrics reflect the characteristics of undervalued stocks, typically calculated through price-to-earnings ratios, price-to-book ratios, or dividend yields. Value metrics support classic value investing principles by capturing market deviations from a company's fundamentals.

[0209] Momentum indicators measure a stock's relative performance over a period of time, typically based on a six- or 12-month cumulative return. Research shows that stocks that have performed well in the past are likely to continue to outperform the market in the short term.

[0210] Volatility: This indicator reflects the historical fluctuations in stock returns. Low-volatility stocks often exhibit a "low volatility anomaly," meaning their risk-adjusted returns outperform those of high-volatility stocks.

[0211] Growth metrics: A measure of a company's growth potential, typically based on revenue or profit growth rates. Growth-metric stocks typically have higher price-to-earnings ratios and are driven by investor preference.

[0212] Profitability: This measure measures the quality and stability of a company's earnings, often expressed as return on equity (ROE) or profit margin. Companies with stable and strong earnings tend to have higher stock returns.

[0213] Liquidity indicator: A measure of the level of stock trading activity, typically based on trading volume or transaction value. Stocks with low liquidity may have a higher liquidity risk premium.

[0214] Leverage indicators reflect a company's level of financial leverage, typically measured by indicators such as the debt-to-asset ratio. Highly leveraged companies may face a higher risk of financial distress.

[0215] Reversal indicators capture the "mean reversion" effect of asset prices, reflecting the tendency of assets to return to their mean after significant fluctuations. Reversal indicators are often used to identify short-term or long-term rebound opportunities.

[0216] This embodiment is based on the entire stock pool of the A-share market and constructs a quantitative stock selection strategy based on indicator domains.

[0217] (1) Data source and processing

[0218] Data source: A-share market data from 2014 to 2024 was obtained from the Wind financial database (as shown in Table 1).

[0219] Data processing: Acquire and pre-process data to ensure it is updated daily and does not contain future information to avoid data leakage.

[0220] Table 1: Example of initial stock data for 2014.

[0221]

[0222] (2) Construction of regional stock pools

[0223] The first selected indicator is obtained as the small market capitalization indicator, and the second time series data set is segmented.

[0224] Segmentation standard: Select the 10% with the lowest market capitalization in the entire market as the small-cap segmentation stock pool.

[0225] (3) Based on a second indicator set that is divided into different categories from the first selected indicator, an indicator analysis and calculation is performed on the second time series data set within the first time interval to obtain the first indicator analysis and calculation result.

[0226] Strategy development: Quantitative strategy development is carried out using the domain analysis module, and evaluation indicators are calculated according to formulas (1.8)-(1.15). Figure 9 The backtest return performance of the entire regional stock as a holding stock from 2014 to 2024 is shown. Figure 9 It can be seen that the net asset value of the regional strategy is better than the global benchmark net asset value. The specific return evaluation indicators are shown in Table 2.

[0227] Table 2: Backtest performance of stocks by domain.

[0228]

[0229] The second indicator set uses style indicators. Since the regional stock pool is determined by the size indicator, the size indicator is excluded from the analysis of the intra-region style indicators to ensure the independence of the results. The intra-region style indicator RankIC matrix is calculated based on formula (1.7).

[0230] Table 3 shows the comparison of IC calculated for style indicators within the domain (small-cap domain) and for the entire stock market.

[0231] Table 3: Comparison of IC calculated for style indicators within the domain and for the entire stock market.

[0232]

[0233] As shown in Table 3, the Reversal Indicator performs best within the IC domain and exhibits the largest absolute change in IC (0.03), significantly outperforming other style indicators. This result suggests that the Reversal Indicator exhibits stronger excess return potential within this small-cap market. Therefore, the Reversal Indicator is selected as the second-choice indicator, representing the optimized strategy indicator.

[0234] Determine the second indicator analysis and calculation results of the individuals in the second time series data set based on the public data information.

[0235] (4) Determine the fourth selection indicator

[0236] Select any common indicator from the reversal indicator library, such as the retracement indicator. Its calculation logic is consistent with formula (1.13). The retracement indicator_60 indicates that the indicator value is the maximum retracement in the past 60 days.

[0237] Perform indicator analysis and calculation on this indicator, and use formulas (1.1)-(1.5) to calculate the RankIC vector of this indicator: .

[0238] The RankIC value here is -0.036, indicating that the backtest indicator has the ability to predict negative returns for all stocks in the market, and the overall prediction success rate is 58.33%.

[0239] Style classification: Calculate the index style exposure vector according to formula (6), and the correlation coefficient is shown in Table 4:

[0240]

[0241] As can be seen from Table 4, the correlation coefficient (0.6) between the retracement indicator and the reversal indicator is significantly higher than that of other style indicators, proving that the retracement indicator (fourth choice indicator) can be classified as a reversal indicator (second choice indicator).

[0242] (5) Calculate the time series change value of the second time series data set based on the first selection indicator and the second selection indicator, and conduct a comparative analysis of the strategy returns.

[0243] The strategy is constructed using the Drawdown Reversal Indicator within the pool of stocks after segmentation. Based on the negative RankIC of the Drawdown Indicator, the 10 stocks with the smallest drawdowns are selected as holdings at the weekly position rotation point during the period of 2014-2024. The final backtest yield curve is as follows: Figure 10 As shown, from Figure 10 It can be seen that the net value of the optimized strategy is significantly better than the net value of the global benchmark, and has been greatly improved compared with the domain strategy using a single selection indicator.

[0244] To verify the effectiveness of the optimized strategy, Table 5 compares the original domain strategy (small market capitalization) with the optimized strategy (small market capitalization + reversal) based on the backtest reversal indicator.

[0245] Table 5: Comparison of original strategy and optimized strategy

[0246]

[0247] As shown in Table 5, the optimized strategy demonstrated outstanding performance in terms of cumulative net value (increased by 785.95%) and annualized return (increased by 48.91%). The maximum drawdown was reduced by 6.58%, further enhancing the strategy's robustness. The optimized strategy's periodic win rate was essentially the same as the original strategy, demonstrating that the improved returns did not sacrifice stability.

[0248] The disclosed embodiments have broad applicability, meeting diverse stock pool analysis needs. They are applicable not only to market-wide stock pools, such as common indices (e.g., the CSI 300 and CSI 1000), but also to stock pools segmented by style indicators. This allows for in-depth analysis of the performance of different indicators within specific sub-domains, further optimizing strategy development.

[0249] The disclosed embodiments are characterized by significant strategy optimization effects. For example, in the disclosed embodiments, the small-cap stock domain strategy carries significant risk. Through the domain analysis module, it was discovered that the reversal indicator performs well in this domain. Based on this discovery, a strategy was constructed using the reversal indicator and compared with the original strategy. The optimized strategy achieved significant improvements in total net value, annualized return, and maximum drawdown. This result verifies the effectiveness of the method of the present invention in improving strategy returns and controlling risks.

[0250] The disclosed embodiments feature efficient data import and indicator calculation. They support operation on local computer platforms, with data reading and processing options available in linear or parallel modes. This balances computational efficiency and operational convenience, providing individual quantitative investors with an easily implemented and reproducible backtesting framework.

[0251] The disclosed embodiment avoids the inability of a single indicator strategy to quickly screen out desired individual stocks. It conducts scientific analysis and evaluation based on the changing patterns of time series data, significantly improving the stock selection efficiency of quantitative investment strategies in specific stock pools and the scientific nature of strategy backtesting, and providing investors with more accurate decision-making support tools.

[0252] The disclosed embodiments can be modularly designed to improve the efficiency and robustness of strategy development. A quantitative stock selection strategy based on indicator domain division is proposed, covering data acquisition and processing, indicator analysis, domain stock analysis, strategy return backtesting and strategy evaluation. The various parts are independent of each other and tightly coupled, which not only improves the efficiency of strategy development, but also facilitates the improvement and optimization of each module, and enhances the robustness and generalization ability of the system. The system adopts a modular design, which improves the flexibility of the system and can be flexibly combined and applied according to different strategy requirements, thereby meeting a variety of backtesting scenarios.

[0253] Figure 11 A block diagram of an electronic device suitable for implementing a time series data selection method based on domain division according to an embodiment of the present disclosure is schematically shown.

[0254] like Figure 11 As shown, the electronic device 1100 according to an embodiment of the present disclosure includes a processor 1101, which can perform various appropriate actions and processes according to a program stored in a read-only memory (ROM) 1102 or a program loaded from a storage portion 1108 into a random access memory (RAM) 1103. The processor 1101 may include, for example, a general-purpose microprocessor (e.g., a CPU), an instruction set processor and / or a related chipset and / or a special-purpose microprocessor (e.g., an application-specific integrated circuit (ASIC)), etc. The processor 1101 may also include onboard memory for caching purposes. The processor 1101 may include a single processing unit or multiple processing units for performing different actions of the method flow according to an embodiment of the present disclosure.

[0255] Various programs and data required for the operation of the electronic device 1100 are stored in the RAM 1103. The processor 1101, the ROM 1102, and the RAM 1103 are connected to each other via a bus 1104. The processor 1101 performs various operations of the method flow according to the embodiment of the present disclosure by executing the programs in the ROM 1102 and / or the RAM 1103. It should be noted that the programs may also be stored in one or more memories other than the ROM 1102 and the RAM 1103. The processor 1101 may also perform various operations of the method flow according to the embodiment of the present disclosure by executing the programs stored in the one or more memories.

[0256] According to an embodiment of the present disclosure, electronic device 1100 may further include an input / output (I / O) interface 1105, which is also connected to bus 1104. Electronic device 1100 may also include one or more of the following components connected to I / O interface 1105: an input section 1106 including a keyboard, mouse, etc.; an output section 1107 including devices such as a cathode ray tube (CRT), liquid crystal display (LCD), and speakers; a storage section 1108 including a hard disk; and a communication section 1109 including a network interface card such as a LAN card or modem. Communication section 1109 performs communication processing via a network such as the Internet. A drive 1110 is also connected to I / O interface 1105 as needed. Removable media 1111, such as a magnetic disk, optical disk, magneto-optical disk, semiconductor memory, etc., is installed in drive 1110 as needed, so that computer programs read from the removable media can be installed into storage section 1108 as needed.

[0257] The present disclosure also provides a computer-readable storage medium, which may be included in the device / apparatus / system described in the above embodiments, or may exist independently and not be incorporated into the device / apparatus / system. The computer-readable storage medium carries one or more programs, and when executed, implements the method according to the embodiments of the present disclosure.

[0258] According to an embodiment of the present disclosure, a computer-readable storage medium may be a non-volatile computer-readable storage medium, and may include, for example, but not limited to: a portable computer disk, a hard disk, a random access memory (RAM), a read-only memory (ROM), an erasable programmable read-only memory (EPROM or flash memory), a portable compact disk read-only memory (CD-ROM), an optical storage device, a magnetic storage device, or any suitable combination thereof. In the present disclosure, a computer-readable storage medium may be any tangible medium that contains or stores a program that can be used by or in conjunction with an instruction execution system, apparatus, or device. For example, according to an embodiment of the present disclosure, a computer-readable storage medium may include the ROM 1102 and / or RAM 1103 described above, and / or one or more memories other than ROM 1102 and RAM 1103.

[0259] The embodiments of the present disclosure also include a computer program product, which includes a computer program containing program code for executing the method shown in the flowchart. When the computer program product is run in a computer system, the program code is used to enable the computer system to implement the method provided by the embodiments of the present disclosure.

[0260] The computer program executes the above functions defined in the system / device of the embodiment of the present disclosure when the computer program is executed by the processor 1101. According to the embodiment of the present disclosure, the system, device, module, unit, etc. described above can be implemented by a computer program module.

[0261] In one embodiment, the computer program may be stored on a tangible storage medium such as an optical storage device or a magnetic storage device. In another embodiment, the computer program may be transmitted and distributed in the form of a signal on a network medium, downloaded and installed via the communication portion 1109, and / or installed from removable media 1111. The program code contained in the computer program may be transmitted using any appropriate network medium, including but not limited to wireless, wired, or any suitable combination thereof.

[0262] In such an embodiment, the computer program can be downloaded and installed from a network via the communication section 1109 and / or installed from the removable medium 1111. When the computer program is executed by the processor 1101, the above-described functions defined in the system of the embodiment of the present disclosure are performed. According to the embodiment of the present disclosure, the systems, devices, means, modules, units, etc. described above can be implemented by computer program modules.

[0263] According to an embodiment of the present disclosure, the program code for executing the computer program provided by the embodiment of the present disclosure can be written in any combination of one or more programming languages. Specifically, these computer programs can be implemented using high-level procedural and / or object-oriented programming languages, and / or assembly / machine languages. Programming languages include, but are not limited to, languages such as Java, C++, Python, "C" or similar programming languages. The program code can be executed entirely on the user computing device, partially on the user device, partially on a remote computing device, or entirely on a remote computing device or server. In cases involving a remote computing device, the remote computing device can be connected to the user computing device through any type of network, including a local area network (LAN) or a wide area network (WAN), or can be connected to an external computing device (for example, using an Internet service provider to connect via the Internet).

[0264] The flowcharts and block diagrams in the accompanying drawings illustrate the possible implementation architecture, functions and operations of the systems, methods and computer program products according to various embodiments of the present disclosure. In this regard, each box in the flowchart or block diagram can represent a module, program segment, or a part of code, and the above-mentioned module, program segment, or a part of code contains one or more executable instructions for implementing the specified logical function. It should also be noted that in some alternative implementations, the functions marked in the box can also occur in an order different from that marked in the accompanying drawings. For example, two boxes represented in succession can actually be executed substantially in parallel, and they can sometimes be executed in the opposite order, depending on the functions involved. It should also be noted that each box in the block diagram or flowchart, and the combination of boxes in the block diagram or flowchart, can be implemented with a dedicated hardware-based system that performs the specified function or operation, or can be implemented with a combination of dedicated hardware and computer instructions.

[0265] Those skilled in the art will appreciate that the features described in the various embodiments of the present disclosure may be combined and / or coupled in various ways, even if such combinations or couplings are not explicitly described in the present disclosure. In particular, the features described in the various embodiments of the present disclosure may be combined and / or coupled in various ways without departing from the spirit and teachings of the present disclosure. All such combinations and / or couplings fall within the scope of the present disclosure.

[0266] The embodiments of the present disclosure are described above. However, these embodiments are for illustrative purposes only and are not intended to limit the scope of the present disclosure. Although each embodiment has been described separately above, this does not mean that the measures in each embodiment cannot be used in combination to advantage. The scope of the present disclosure is defined by the appended claims and their equivalents. Without departing from the scope of the present disclosure, those skilled in the art may make various substitutions and modifications, which should all fall within the scope of the present disclosure.

Claims

1. A domain-based indicator selection strategy construction system, characterized by: Include: a data import module configured to, in response to obtaining a first selection index from a client, call a first data interface based on the first selection index to determine and import a second time series data set within a first time interval from the first time series data set; an indicator calculation module, configured to perform indicator calculation on the second time series data set based on the first selected indicator and a second indicator set divided into a different category from the first selected indicator, to obtain a first indicator calculation result; an indicator analysis module, configured to perform an indicator analysis calculation based on the first indicator calculation result to obtain a first indicator analysis result; The domain analysis module is configured to determine a second selection indicator associated with the first selection indicator from the second indicator set based on the first indicator analysis result and the second time series data set.

2. The system according to claim 1, wherein: The system further comprises: A second data interface is configured to crawl public data information related to individuals in the second time series dataset within the first time interval from public data according to a call; an information analysis module, configured to determine a public data analysis result of an individual of the second time series dataset based on the public data information; The indicator analysis module is further configured to perform the indicator analysis calculation based on the public data analysis result to obtain a second indicator analysis result; The domain analysis module is further configured to determine a third selection indicator associated with the first selection indicator based on the second indicator analysis result.

3. The system according to claim 1, wherein: The indicator calculation module is further configured to perform indicator analysis and calculation on the first time series data set based on a fourth selection indicator set associated with the second selection indicator to obtain a third indicator analysis and calculation result; The domain analysis module is further configured to determine, based on the calculation result of the third indicator analysis, a fourth selection indicator associated with the first selection indicator from the fourth selection indicator set.

4. The system according to claim 1, wherein: Also includes: The strategy building module is configured to determine the individuals of the second time series data set based on the first selection index and the second selection index.

5. A method for constructing an indicator selection strategy based on domain division, characterized in that: include: In response to obtaining a first selection indicator from the client, the data import module calls a first data interface based on the first selection indicator to determine and import a second time series data set within a first time interval from the first time series data set; An indicator calculation module performs a first indicator calculation on the second time series data set based on the first selection indicator and a second indicator set divided into different categories of the first selection indicator to obtain a first indicator calculation result; Importing the first indicator calculation result into an indicator analysis module to perform indicator analysis calculation to obtain a first indicator analysis result; The domain analysis module determines a second selection indicator associated with the first selection indicator from the second indicator set based on the first indicator analysis result and the second time series data set.

6. The method according to claim 5, characterized in that Also includes: Invoke a second data interface to crawl public data information related to individuals in the second time series dataset within the first time interval from public data; The information analysis module determines a public data analysis result of the individual of the second time series dataset based on the public data information; Importing the public data analysis result into the index analysis module to perform index analysis calculation to obtain a second index analysis result; The domain analysis module determines a third selection indicator associated with the first selection indicator based on the second indicator analysis result.

7. The method according to claim 5, characterized in that Also includes: The strategy construction module determines the individuals of the second time series data set based on the first selection index and the second selection index.

8. An electronic device comprising: one or more processors; a memory for storing one or more computer programs, It is characterized in that the one or more processors execute the one or more computer programs to implement the steps of the method according to any one of claims 5 to 7.

9. A computer-readable storage medium having a computer program stored thereon, characterized in that: When the computer program is executed by a processor, the steps of the method according to any one of claims 5 to 7 are implemented.

10. A computer program product comprising a computer program, characterized in that When the computer program is executed by a processor, the steps of the method according to any one of claims 5 to 7 are implemented.

Images