WO2020087791A1 - 基金信息的分析方法及装置、存储介质、计算机设备 - Google Patents

基金信息的分析方法及装置、存储介质、计算机设备 Download PDF

Info

Publication number
WO2020087791A1
WO2020087791A1 PCT/CN2019/073288 CN2019073288W WO2020087791A1 WO 2020087791 A1 WO2020087791 A1 WO 2020087791A1 CN 2019073288 W CN2019073288 W CN 2019073288W WO 2020087791 A1 WO2020087791 A1 WO 2020087791A1
Authority
WO
WIPO (PCT)
Prior art keywords
stock
fund
information
funds
preset
Prior art date
Legal status (The legal status is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the status listed.)
Ceased
Application number
PCT/CN2019/073288
Other languages
English (en)
French (fr)
Inventor
李海疆
Current Assignee (The listed assignees may be inaccurate. Google has not performed a legal analysis and makes no representation or warranty as to the accuracy of the list.)
Ping An Technology Shenzhen Co Ltd
Original Assignee
Ping An Technology Shenzhen Co Ltd
Priority date (The priority date is an assumption and is not a legal conclusion. Google has not performed a legal analysis and makes no representation as to the accuracy of the date listed.)
Filing date
Publication date
Application filed by Ping An Technology Shenzhen Co Ltd filed Critical Ping An Technology Shenzhen Co Ltd
Publication of WO2020087791A1 publication Critical patent/WO2020087791A1/zh
Anticipated expiration legal-status Critical
Ceased legal-status Critical Current

Links

Images

Classifications

    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/06—Asset management; Financial planning or analysis
    • G—PHYSICS
    • G06—COMPUTING OR CALCULATING; COUNTING
    • G06Q—INFORMATION AND COMMUNICATION TECHNOLOGY [ICT] SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES; SYSTEMS OR METHODS SPECIALLY ADAPTED FOR ADMINISTRATIVE, COMMERCIAL, FINANCIAL, MANAGERIAL OR SUPERVISORY PURPOSES, NOT OTHERWISE PROVIDED FOR
    • G06Q40/00—Finance; Insurance; Tax strategies; Processing of corporate or income taxes
    • G06Q40/04—Trading; Exchange, e.g. stocks, commodities, derivatives or currency exchange

Definitions

  • This application relates to the field of data analysis technology, and in particular, to a method and device for analyzing fund information, storage media, and computer equipment.
  • Stock funds also known as stock funds, refer to funds that invest in the stock market. There are many types of securities funds. Investment strategies are value-based, growth-oriented, and balanced. There are as many as 246 closed-end stock funds. The strategies of these funds are similar and different. Therefore, companies or individuals need to conduct professional analysis on the selection of stock funds to avoid risks as much as possible.
  • the analysis of stock funds mainly uses conventional analysis methods, that is, statistics of several stock funds with good recent returns, and then analyzes the stock investment strategies of these stock funds.
  • conventional analysis methods that is, statistics of several stock funds with good recent returns, and then analyzes the stock investment strategies of these stock funds.
  • the current conventional analysis method cannot find relatively effective stock investment strategies and stable performance stock funds in the market.
  • this application provides a fund information analysis method and device, storage medium, and computer equipment.
  • the main purpose is to solve the conventional analysis method of using stock funds, unable to analyze the relatively effective stock investment strategies and performance in the market The problem of stable stock funds.
  • a method for analyzing fund information includes:
  • the contribution degree of each stock type fund to the deviation feature vector information is analyzed to obtain a target stock type fund that meets the preset stable performance conditions.
  • a device for analyzing fund information includes:
  • An obtaining unit configured to obtain information data corresponding to each stock fund currently in circulation, and the information data includes daily closing net worth information corresponding to each stock;
  • the processing unit is configured to perform standardization processing according to the daily closing net value information corresponding to each stock fund to obtain logarithmic return sequence matrix information;
  • a calculation unit configured to calculate sample covariance matrix information corresponding to the logarithmic return sequence matrix information
  • a filtering unit configured to filter the deviation feature vector information corresponding to the sample covariance matrix information
  • the analysis unit is used to analyze the contribution degree of each stock type fund to the deviation feature vector information, and obtain a target stock type fund that meets the preset stable performance conditions.
  • a non-volatile readable storage medium on which computer readable instructions are stored, which when executed by a processor implements the above method for analyzing fund information.
  • a computer device including a non-volatile readable storage medium, a processor, and a computer readable stored on the non-volatile readable storage medium and operable on the processor Instructions, when the processor executes the computer-readable instructions, the method for analyzing the fund information described above is implemented.
  • this application provides a fund information analysis method and device, storage medium, and computer equipment.
  • This application The daily closing net value information is standardized to obtain the logarithmic return sequence matrix information, and then the sample covariance matrix information corresponding to the logarithmic return sequence matrix information is deviated from the feature vector information, and finally analyzed by each stock fund.
  • the contribution of the vector information can be used to obtain target stock funds that meet the predetermined performance stability conditions, and then real-time target stock funds with relatively stable and stable returns in the market can be found according to changes in the stock fund market, and according to the target stock funds
  • the investment objectives in the public information are analyzed and the corresponding stock investment strategy is analyzed.
  • FIG. 2 shows a schematic flowchart of another method for analyzing fund information provided by an embodiment of the present application
  • FIG. 3 shows a schematic structural diagram of a fund information analysis device provided by an embodiment of the present application
  • FIG. 4 shows a schematic structural diagram of another fund information analysis device provided by an embodiment of the present application.
  • Methods include:
  • the information data includes the daily closing net worth information corresponding to each stock fund, the equity data of each stock under the fund (that is, the stocks that can be circulated in the stock exchange market), and the suspension and resumption data (the data of suspended and resumed stocks) ), Company behavior data (data on the distribution of company stock dividends), etc.
  • the daily closing net value information refers to the daily net asset value of the fund unit, which is an important indicator reflecting the performance of the fund.
  • the transaction price of the open-end fund is determined based on the net value of each fund unit.
  • total assets refer to all assets owned by the fund; total liabilities refer to liabilities formed during the operation and financing of the fund, including various fees payable to others, interest payable to funds, etc .; the total number of fund shares refers to the funds issued at the time The total amount of shares.
  • the total number of open-end fund shares changes every day, so the statistics after the end of the day ’s trading must prevail. After the market closes on each business day, divide the net asset value of the day ’s fund by the total number of fund shares at the end of the day ’s trading. Net asset value of the fund unit on the day of issue.
  • the executive body of the present invention may be a device or device for analyzing fund information, which is used to obtain daily closing net value information and logarithmic return sequence matrix information in the information data corresponding to each stock type fund currently in circulation, and is also used to calculate Sample covariance matrix information, filter sample covariance matrix information corresponding to the deviation feature vector information, analyze the contribution of each stock fund to the deviation feature vector information, and obtain the stock investment strategy information corresponding to the target stock fund, specifically perform the following steps The process shown.
  • the standardization process is to scale the data so that it falls into a small specific interval.
  • linear method such as extreme value method, standard deviation method
  • polyline method such as trifold method
  • curvilinear method such as semi-normal distribution
  • the sample covariance information is the sample covariance matrix, which is a matrix constructed with multidimensional random samples, which is usually used as an estimate of the population covariance matrix D (X), and the sample covariance matrix is the population covariance Unbiased estimate of matrix D (X).
  • the deviation feature vector is the feature vector corresponding to the abnormal feature value, that is, the feature vector corresponding to the feature value of the sample covariance matrix.
  • this solution is to obtain information data corresponding to each stock fund currently in circulation and use the daily data contained in the consultation data Closing the net value information to obtain the logarithmic return sequence matrix information, sequentially calculating the sample covariance matrix information, filtering out the deviation feature vector information, and finally by analyzing the contribution of each stock fund to the deviation feature vector information respectively, to obtain the preset performance
  • the target stock fund with stable conditions can obtain the stock investment strategy information corresponding to the target stock fund. Make the analysis results have basis, and can obtain the corresponding results in real time according to the market changes of the stock fund, which is very convenient and intelligent, and can provide important reference resources for enterprises or individuals to choose stock funds.
  • the method includes:
  • the information data corresponding to stock fund a, stock fund b, and stock fund c are respectively obtained from the information terminal, and the daily closing net value corresponding to each stock fund and each fund under the fund can be obtained according to the information data.
  • Data information such as stock equity data, suspension and resumption data, and company behavior data.
  • the second preset formula is Is the net closing value of stock i on day t, Is the net closing value of stock i on day t-1, Is the logarithmic return of stock i on day t, T is the number of observation points, ⁇ i is the standard deviation of the logarithmic return of stock i, and N is the number of stocks.
  • the expression form of the above formula is only an optimal formula given in the embodiment of the present invention, and the logarithmic return sequence matrix A can also be calculated based on the above formula, such as transforming letters in the formula, adding corrections Coefficients and the like are not limited in the embodiments of the present invention.
  • the third preset formula is A ′ is the transposed matrix of A, and ⁇ N * N is the sample covariance matrix.
  • the filter matrix can be calculated after the sample covariance matrix is obtained It is used to denoise the covariance matrix, because Markwitz's mean-variance model uses the overall covariance matrix to describe risk, but in reality, it is not known what the overall covariance is, so the way people take is to use
  • the sample covariance matrix is used to approximate the description.
  • the elements in the covariance matrix are calculated using historical data (such as fund value data for consecutive T days). According to the statistical theorem, the sample covariance matrix only reaches positive infinity when T Time is the unbiased estimate of the overall covariance matrix.
  • the sample covariance matrix calculated using the limited historical data is a biased estimate of the overall covariance matrix, and the degree of this deviation will increase with the increase of the matrix dimension.
  • Dimension disaster ".
  • the covariance matrix contains a lot of white noise.
  • the expression form of the above formula is only an optimal formula given in the embodiment of the present invention, and it can also be deformed based on the above formula to calculate each eigenvalue of the sample covariance matrix and the corresponding eigenvector of each eigenvalue.
  • the letters in the transformation formula, the increase of the weighting coefficient, and the addition of the correction coefficient are not limited in this embodiment of the present invention.
  • the fourth preset formula is ⁇ max is the theoretical maximum boundary value, which can be obtained according to the MP law (Marcenko-Pastur Law).
  • the MP theorem the specific content refers to, for a large-dimensional random matrix, if all matrix elements are independent and identically distributed, then the matrix The density function of the empirical distribution function of the eigenvalues satisfies a specific law.
  • the MP law is the semicircle rate. Q is obtained by dividing the number of stock funds by T.
  • the abnormal feature value bit feature value does not match the feature value of the ball.
  • a fund code “150008” the fund name is “SDIC UBS Ruihe Xiaokang”
  • the fund strategy is "the fund through passive index investment management, to achieve effective tracking of the Shanghai and Shenzhen 300 index, and strive to fund
  • the daily average tracking error between the rate of return and the benchmark of performance should be controlled within 0.35%, and the annual tracking error should be controlled within 4%.
  • the fund issuing company is "SDB UBS";
  • Each stock fund is classified according to the fund issuing company. Since a and b are two stock funds of the same fund issuing company, a and b can be classified into one category, and for subsequent processing, the same classification code is configured , Such as "13".
  • the first preset formula is Is the contribution degree of the stock fund whose classification code is s to the k -th largest deviation feature vector u (k) in the deviation feature vector information; if the index code of the index fund i is s, then n s represents the number of stock funds classified as s, if the index code of index fund i is not s, then Is the ith component of u (k) .
  • the expression form of the above formula is only an optimal formula given in the embodiment of the present invention, and the contribution degree of each stock fund to the deviation feature vector information can also be calculated based on the above formula.
  • the letters in the transformation formula, the increase of the weighting coefficient, and the addition of the correction coefficient are not limited in this embodiment of the present invention.
  • step 209 specifically includes: determining the preset number of stock funds with the highest contribution degree as the target stock fund that meets the preset stable performance conditions; or statistics The predetermined number of target fund issuing companies with the highest contribution rank; the stock funds currently issued by the predetermined number of target fund issuing companies respectively are determined as the target stock fund that meets the preset stable performance conditions. For example, if the target stock fund is determined according to the order of stock funds, and the preset number of target stock funds to be extracted is 10, then the top 10 stock funds will be selected as targets that meet the preset stable performance conditions according to the contribution ranking Stock funds.
  • the target stock fund is determined according to the order of fund issuing companies, and the preset number of target fund issuing companies to be extracted is 10, the top 10 fund issuing companies in the contribution contribution ranking will be extracted.
  • the issued stock fund is determined as the target stock fund that meets the preset conditions for stable performance.
  • the correlation coefficient When the correlation coefficient is less than 0, it is called negative correlation; when it is greater than 0, it is called positive correlation; when it is equal to 0 , Called zero correlation.
  • fund A uses a passive investment strategy that tracks GEM
  • fund B uses a passive investment strategy that tracks SSE 50.
  • the correlation coefficient of the two funds in this time period is -0.8, then, It can be concluded that the strategy of fund A and the strategy of fund B are negatively correlated.
  • funds in the fund (Fund of Funds, FOF)
  • funds A and B can be held at the same time to achieve risk hedging; if After analyzing that funds A and B are stable funds in the market, it can be determined that the correlation between the two is relatively reliable.
  • the above method can further determine the internal contact information between the stock investment strategy information of the target stock fund and the stock investment strategy information of other stock funds, so as to obtain more stock investment strategy information, and analyze the Correlation, in turn, can extract the market performance and collocation effects between different strategies, providing more possibilities for companies or individuals to choose in stock funds.
  • an embodiment of the present application provides a fund information analysis device.
  • the device includes: an acquisition unit 31, a processing unit 32, and a calculation unit 33. Screening unit 34, analysis unit 35.
  • the obtaining unit 31 can be used to obtain information data corresponding to each stock fund currently in circulation, and the information data includes daily closing net worth information corresponding to each stock;
  • the processing unit 32 can be used for standardization processing according to the daily closing net value information corresponding to each stock fund to obtain logarithmic return sequence matrix information;
  • the calculation unit 33 can be used to calculate the sample covariance matrix information corresponding to the logarithmic return sequence matrix information
  • the filtering unit 34 can be used to filter the deviation feature vector information corresponding to the sample covariance matrix information
  • the analysis unit 35 can be specifically used to classify each stock fund according to the fund issuing company and configure the classification code corresponding to each fund issuing company; Calculate the contribution degree of each stock fund to the deviation feature vector information, where, Is the contribution degree of the stock fund whose classification code is s to the k -th largest deviation feature vector u (k) in the deviation feature vector information; if the index code of the index fund i is s, then n s represents the number of stock funds classified as s, if the index code of index fund i is not s, then Is the i-th component of u (k) ; rank the contribution degree in descending order, and according to the preset number of stock funds with the highest contribution degree, determine the target stocks that meet the preset stable performance conditions Fund.
  • the analysis unit 35 can also be specifically used to determine the preset number of stock funds as the target stock funds that meet the preset stable performance conditions; or the predetermined number of top ranked statistical contributions Target fund issuing company; determine the stock funds currently issued separately by a predetermined number of target fund issuing companies as target stock funds that meet the preset conditions for stable performance.
  • the processing unit 32 may be specifically used to determine the daily closing net value of each stock fund in the daily closing net value information as a parameter, according to the second preset formula Calculate the logarithmic yield sequence matrix A;
  • calculation unit 33 can be specifically used to follow the third preset formula Calculate the sample covariance matrix ⁇ N * N corresponding to the logarithmic yield sequence matrix A, A ′ is the transposed matrix of A.
  • the screening unit 34 can be specifically used to calculate each eigenvalue of the sample covariance matrix ⁇ N * N and the corresponding eigenvalue of each eigenvalue; if the elements in the sample covariance matrix ⁇ N * N meet the independent Same distribution, then use the fourth preset formula Calculate the theoretical maximum boundary value of the eigenvalues of the sample covariance matrix ⁇ N * N , ⁇ max is the theoretical maximum boundary value, Q is obtained by dividing the number of stock funds by T; determine the eigenvalue of each eigenvalue greater than ⁇ max Is the abnormal feature value, and the feature vector corresponding to the abnormal feature value is determined as the deviation feature vector corresponding to the sample covariance matrix information.
  • the device may also include: a determination unit 36.
  • the calculation unit 33 can specifically be used to calculate the correlation information between the target stock fund and other stock funds; the determination unit 36 can be used to determine the stock investment strategy information of the target stock fund and other stock types based on the correlation information Intrinsic contact information between the fund's stock investment strategy information.
  • this embodiment also provides a non-volatile readable storage medium on which computer-readable instructions are stored, and when the readable instructions are executed by the processor.
  • the above-mentioned method for analyzing fund information shown in FIGS. 1 to 2 is implemented.
  • the technical solution of the present application can be embodied in the form of a software product, which can be stored in a non-volatile storage medium (can be a CD-ROM, U disk, mobile hard disk, etc.), including several The instructions are used to enable a computer device (which may be a personal computer, a server, or a network device, etc.) to execute the methods described in various implementation scenarios of the present application.
  • a computer device which may be a personal computer, a server, or a network device, etc.
  • this embodiment also provides a computer device, the computer device includes a non-volatile Readable storage medium and processor; non-volatile readable storage medium for storing computer-readable instructions; processor for executing computer-readable instructions to implement the above-mentioned fund information shown in FIGS. 1 to 2 Analytical method.
  • the computer device may further include a user interface, a network interface, a camera, a radio frequency (Radio Frequency) circuit, a sensor, an audio circuit, a WI-FI module, and so on.
  • the user interface may include a display (Display), an input unit such as a keyboard, and the like, and the optional user interface may also include a USB interface, a card reader interface, and the like.
  • the network interface may optionally include a standard wired interface, a wireless interface (such as a WI-FI interface), and so on.
  • the computer device structure provided in this embodiment does not constitute a limitation on the physical device, and may include more or less components, or combine some components, or arrange different components.
  • the non-volatile storage medium may also include an operating system and a network communication module.
  • An operating system is a computer-readable instruction that manages the hardware and software resources of the above-mentioned computer equipment, and supports the operation of information-processing computer-readable instructions and other software and / or computer-readable instructions.
  • the network communication module is used to realize communication between various components inside the non-volatile storage medium, and to communicate with other hardware and software in the information processing entity device.
  • the target stock fund can also calculate the correlation information between the target stock fund and other stock funds, determine the internal contact information between the target stock fund's stock investment strategy information and other stock fund's stock investment strategy information, and then be able to
  • the stock investment strategy information is enriched and expanded to extract the market performance and matching effect between different strategies.

Landscapes

  • Business, Economics & Management (AREA)
  • Engineering & Computer Science (AREA)
  • Finance (AREA)
  • Accounting & Taxation (AREA)
  • Development Economics (AREA)
  • Theoretical Computer Science (AREA)
  • Physics & Mathematics (AREA)
  • General Physics & Mathematics (AREA)
  • General Business, Economics & Management (AREA)
  • Economics (AREA)
  • Marketing (AREA)
  • Strategic Management (AREA)
  • Technology Law (AREA)
  • Entrepreneurship & Innovation (AREA)
  • Operations Research (AREA)
  • Human Resources & Organizations (AREA)
  • Game Theory and Decision Science (AREA)
  • Financial Or Insurance-Related Operations Such As Payment And Settlement (AREA)

Abstract

本申请公开了一种基金信息的分析方法及装置、存储介质、计算机设备，涉及数据分析技术领域，可以解决无法科学分析出股票型基金市场上有效的股票投资策略和稳定的股票型基金的问题。其中方法包括：获取当前流通的各只股票型基金对应的资讯数据，资讯数据中包含各只股票型基金对应的每日收盘净值信息；按照各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息；计算对数收益率序列矩阵信息对应的样本协方差矩阵信息；筛选样本协方差矩阵信息对应的偏离特征向量信息；分析各个股票型基金分别对偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金，以便获取目标股票型基金对应的股票投资策略信息。

Description

基金信息的分析方法及装置、存储介质、计算机设备
本申请要求与2018年10月30日提交中国专利局、申请号为2018112804488、申请名称为“基金信息的分析方法及装置、存储介质、计算机设备”的中国专利申请的优先权，其全部内容通过引用结合在申请中。
技术领域
本申请涉及数据分析技术领域，尤其涉及到一种基金信息的分析方法及装置、存储介质、计算机设备。
背景技术
股票型基金又称股票基金，是指投资于股票市场的基金。证券基金的种类很多。投资策略有价值型、成长型、平衡型。封闭式股票型基金多达246只，这些基金的策略有相近也有不同。因此企业或个人在股票型基金的选择上需要进行专业的分析，尽量规避风险。
目前股票型基金的分析主要是采用常规分析方法，即统计近期收益较好的几个股票型基金，然后再分析这几个股票型基金的股票投资策略。然而，由于股票市场变动瞬息万变，面对市场上数量众多的股票型基金，利用目前的常规分析法无法找出市场上相对有效的股票投资策略和表现稳定的股票型基金。
发明内容
有鉴于此，本申请提供了一种基金信息的分析方法及装置、存储介质、计算机设备，主要目的在于解决利用股票型基金的常规分析方法，无法分析出市场上相对有效的股票投资策略和表现稳定的股票型基金的问题。
根据本申请的一个方面，提供了一种基金信息的分析方法，该方法包括：
获取当前流通的各只股票型基金对应的资讯数据，所述资讯数据中包含所述各只股票对应的每日收盘净值信息；
按照所述各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息；
计算所述对数收益率序列矩阵信息对应的样本协方差矩阵信息；
筛选所述样本协方差矩阵信息对应的偏离特征向量信息；
分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金。
根据本申请的另一个方面，提供了一种基金信息的分析装置，该装置包括：
获取单元，用于获取当前流通的各只股票型基金对应的资讯数据，所述资讯数据中包含所述各只 股票对应的每日收盘净值信息；
处理单元，用于按照所述各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息；
计算单元，用于计算所述对数收益率序列矩阵信息对应的样本协方差矩阵信息；
筛选单元，用于筛选所述样本协方差矩阵信息对应的偏离特征向量信息；
分析单元，用于分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金。
根据本申请的又一个方面，提供了一种非易失性可读存储介质，其上存储有计算机可读指令，所述计算机可读指令被处理器执行时实现上述基金信息的分析方法。
根据本申请的再一个方面，提供了一种计算机设备，包括非易失性可读存储介质、处理器及存储在非易失性可读存储介质上并可在处理器上运行的计算机可读指令，所述处理器执行所述计算机可读指令时实现上述基金信息的分析方法。
借由上述技术方案，本申请提供的一种基金信息的分析方法及装置、存储介质、计算机设备，与目前股票型基金采用的常规分析方法相比，本申请按照各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息，然后再筛选对数收益率序列矩阵信息对应的样本协方差矩阵信息的偏离特征向量信息，最后分析各个股票型基金分别对偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金，进而可以实时根据股票基金市场的变动找出市场上收益表现相对稳定、持久的目标股票型基金，并根据目标股票型基金公开信息中的投资目标，分析出相应的股票投资策略。通过这种方法，能够有效提高分析结果的科学性，从而为企业或个人在股票型基金的选择上起到专业的参考作用，从而使投资者能最大程度的规避投资风险。
上述说明仅是本申请技术方案的概述，为了能够更清楚了解本申请的技术手段，而可依照说明书的内容予以实施，并且为了本申请的上述和其他目的、特征和优点能够更明显易懂，以下特举本申请的具体实施方式。
附图说明
此处所说明的附图用来提供对本申请的进一步理解，构成本申请的一部分，本申请的示意性实施例及其说明用于解释本申请，并不构成对本地申请的不当限定。在附图中：
图1示出了本申请实施例提供的一种基金信息的分析方法的流程示意图；
图2示出了本申请实施例提供的另一种基金信息的分析方法的流程示意图；
图3示出了本申请实施例提供的一种基金信息的分析装置的结构示意图；
图4示出了本申请实施例提供的另一种基金信息的分析装置的结构示意图。
具体实施方式
下文中将参照附图并结合实施例来详细说明本申请。需要说明的是，在不冲突的情况下，本申请实施例及实施例中的特征可以相互结合。
针对目前无法科学地筛选出整个股票型基金市场上相对有效的股票投资策略和表现稳定的股票型基金的问题，本申请实施例提供了一种基金信息的分析方法，如图1所示，该方法包括：
101、获取当前流通的各只股票型基金对应的资讯数据。
其中，资讯数据中包含各只股票型基金对应的每日收盘净值信息、基金下各只股票的股本数据(即证券交易市场上能够流通的股票)、停复牌数据(停牌股票和复牌股票的数据)、公司行为数据(公司股票红利的分配情况数据)等。每日收盘净值信息是指每日基金单位资产净值，它是反映基金绩效表现的一个重要指标，开放式基金的交易价格就是以每基金单位的净值为依据确定的。每日基金净值计算公式为：基金份额资产净值＝(总资产-总负债)/基金份额总数。其中，总资产是指基金拥有的所有资产；总负债是指基金运作及融资时所形成的负债，包括应付给他人的各项费用、应付资金利息等；基金份额总数是指当时发行在外的基金份额的总量。开放式基金的份额总数每天都在变化，因此须以当日交易结束后的统计数为准，在每个营业日收市后，将当日基金资产净值除以当日交易截止时的基金份额总数，就得出当日的基金单位资产净值。
本发明的执行主体可以为基金信息分析的装置或设备，用于获取当前流通的各只股票型基金对应的资讯数据中的每日收盘净值信息以及对数收益率序列矩阵信息，还用于计算样本协方差矩阵信息、筛选样本协方差矩阵信息对应的偏离特征向量信息，分析各个股票型基金分别对偏离特征向量信息的贡献度，获取目标股票型基金对应的股票投资策略信息，具体执行以下步骤所示过程。
102、按照各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息。
其中，标准化处理是将数据按比例缩放，使之落入一个小的特定区间。目前数据标准化方法有多种，归结起来可以分为直线型方法(如极值法、标准差法)、折线型方法(如三折线法)、曲线型方法(如半正态性分布)。
103、计算对数收益率序列矩阵信息对应的样本协方差矩阵信息。
其中，样本协方差信息即为样本协方差矩阵，它是用多维随机样本构造的矩阵，它通常用来作为总体协方差矩阵D(X)的估计量，而且，样本协方差矩阵是总体协方差矩阵D(X)的无偏估计。
104、筛选样本协方差矩阵信息对应的偏离特征向量信息。
其中，偏离特征向量是异常特征值对应的特征向量，即偏离样本协方差矩阵的特征值对应的特征向量。
105、分析各个股票型基金分别对偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金。
其中，贡献度为是分析经济效益的一个指标，可用于分析经济增长中各因素作用大小的程度。计算公式：贡献率(％)＝x因素贡献量(增量或增长程度)/总贡献量(总增量或增长程度)×100％，指x因素的增长量(程度)占总增长量(程度)的比重。例如，股票型基金a贡献率(％)＝股票型基金a资产额/资产总额×100％。，在确定目标股票型基金后，可获取目标股票型基金对应的股票投资策略信息，该股票投资策略信息是股票投资者为避免或降低风险，获取较多投资收益而采用的方法和措施。
通过本实施例中的基金信息的分析方法，与现有基金信息的分析方法相比，本方案是通过获取当前流通的各只股票型基金对应的资讯数据，并利用咨询数据中包含的每日收盘净值信息得到对数收益率序列矩阵信息，依次计算出样本协方差矩阵信息，筛选出偏离特征向量信息，最后通过分析各个股票型基金分别对偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金，获取到目标股票型基金对应的股票投资策略信息。使分析结果具有依据，并且可以随时根据股票型基金的市场变化，实时得到对应的结果，非常的方便智能，可以为企业或个人对股票型基金的选择上提供重要的参考资源。
进一步的，作为上述实施例具体实施方式的细化和扩展，为了完整说明本实施里中的具体实施过程，提供了另一种基金信息的分析方法，如图2所示，该方法包括：
201、获取当前流通的各只股票型基金对应的资讯数据。
例如，从资讯终端分别获取到股票型基金a、股票型基金b、股票型基金c对应的资讯数据，并可根据资讯数据获取到各只股票型基金对应的每日收盘净值、基金下各只股票的股本数据、停复牌数据、公司行为数据等数据信息。
202、依据每日收盘净值信息中各只股票型基金的每日收盘净值为参数，按照第二预设公式计算对数收益率序列矩阵A。
其中，第二预设公式为
Figure PCTCN2019073288-appb-000001
Figure PCTCN2019073288-appb-000002
是股票i在第t日的收盘净值，
Figure PCTCN2019073288-appb-000003
是股票i在第t-1日的收盘净值，
Figure PCTCN2019073288-appb-000004
是股票i在第t日的对数收益率，
Figure PCTCN2019073288-appb-000005
T为观测时点个数，
Figure PCTCN2019073288-appb-000006
δ i为股票i的对数收益率的标准差，N为股票数量。
需要说明的是，上述公式的表现形式只是本发明实施例中给出的一个最优公式，还可以基于上述公式进行变形来计算对数收益率序列矩阵A，如变换公式中的字母、增加修正系数等，在此本发明实施例不做限定。
203、按照第三预设公式计算对数收益率序列矩阵A对应的样本协方差矩阵。
其中，第三预设公式为
Figure PCTCN2019073288-appb-000007
A′为A的转置矩阵，∑ N*N为样本协方差矩阵。
在具体的实施方式中，在得到样本协方差矩阵后可计算过滤矩阵
Figure PCTCN2019073288-appb-000008
用于给协方差矩阵去噪，因为马克维茨的均值-方差模型是以总体的协方差矩阵来描述风险的，但是现实中，并不知道总体协方差是什么，所以人们采取的办法是利用样本协方差矩阵来近似刻画，具体来说，就是用历史数据(比如连续T天的基金净值数据)计算协方差矩阵里的元素，根据统计学定理，样本协方差矩阵只有在T趋于正无穷大时才是总体协方差阵的无偏估计量。换句话说，由于数据的有限性，利用有限的历史数据计算出来的样本协方差阵是总体协方差阵的有偏估计，而且这种偏离程度会随着矩阵维度的增加而增加，即所谓“维度灾难”。在这种情况下，协方差矩阵是含有大量的白噪音的。
204、计算样本协方差矩阵的各个特征值以及各个特征值相应的特征向量。
其中，计算样本协方差矩阵∑ N*N的各个特征值以及各个特征值相应的特征向量的公式为：
Figure PCTCN2019073288-appb-000009
λ (i)为∑ N*N的升序排列中排名第i位的特征值，I代表单位矩阵，用于对样本协方差进行特征值求解，根据Aμ＝λμ可求解出特征向量μ(i)。
需要说明的是，上述公式的表现形式只是本发明实施例中给出的一个最优公式，还可以基于上述公式进行变形来计算样本协方差矩阵的各个特征值以及各个特征值相应的特征向量。如变换公式中的字母、增加权重系数、增加修正系数等，在此本发明实施例不做限定。
205、若样本协方差矩阵中的元素符合独立同分布，则利用第四预设公式计算样本协方差矩阵的 特征值的理论最大边界值。
其中，第四预设公式为
Figure PCTCN2019073288-appb-000010
λ max为理论最大边界值，可根据M-P定律(Marcenko-Pastur Law)得到，M-P定理，具体内容是指，对于大维随机矩阵而言，如果所有矩阵元素是独立同分布的话，那么这个矩阵的特征值的经验分布函数的密度函数满足特定规律，当对数据进行标准化处理后，M-P定律就是半圆率。而Q由股票型基金的数量除以T得到。
206、将各个特征值中大于理论最大边界值的特征值确定为异常特征值，并将异常特征值相应的特征向量确定为样本协方差矩阵信息对应的偏离特征向量。
其中，异常特征值位特征值中不符合要球的特征值。
例如，若计算出λ (1)大于理论最大边界值λ max，则说明λ (1)为异常特征值，并将异常特征值λ (1)相应的特征向量μ(1)确定为样本协方差矩阵信息对应的偏离特征向量。
207、将各个股票型基金按照基金发行公司进行归类，并配置每个基金发行公司对应的归类编码。
例如，a基金代码“150008”，基金名称为“国投瑞银瑞和小康”，基金策略为“本基金通过被动的指数化投资管理，实现对沪深300指数的有效跟踪，力求将基金净值收益率与业绩比较基准之间的日平均跟踪误差控制在0.35％以内，年跟踪误差控制在4％以内”，基金发行公司为“国投瑞银”；
b基金代码“150009”，基金名称为“国投瑞银瑞和远见”，本基金通过被动的指数化投资管理，实现对沪深300指数的有效跟踪，力求将基金净值收益率与业绩比较基准之间的日平均跟踪误差控制在0.35％以内，年跟踪误差控制在4％以内”，基金发行公司为“国投瑞银”。
将各个股票型基金按照基金发行公司进行归类，因a和b是相同基金发行公司的两个股票基金，故可将a和b归为一类，并且为了后续处理，配置相同的归类代码，如“13”。
208、按照第一预设公式计算各个股票型基金对偏离特征向量信息的贡献度。
其中，第一预设公式为
Figure PCTCN2019073288-appb-000011
Figure PCTCN2019073288-appb-000012
为归类编码为s的股票型基金对偏离特征向量信息中排序第k大的偏离特征向量u (k)的贡献度；若指数型基金i的归类编码为s，则
Figure PCTCN2019073288-appb-000013
n s代表归类编码为s的股票型基金的基金数量，若指数型基金i的归类编码不为s，则
Figure PCTCN2019073288-appb-000014
Figure PCTCN2019073288-appb-000015
为u (k)的第i个分量。
需要说明的是，上述公式的表现形式只是本发明实施例中给出的一个最优公式，还可以基于上述公式进行变形来计算各个股票型基金对偏离特征向量信息的贡献度。如变换公式中的字母、增加权重系数、增加修正系数等，在此本发明实施例不做限定。
209、将贡献度按照从大到小的顺序排序，并依据贡献度排名靠前的预设个数的股票型基金，确定符合预设表现稳定条件的目标股票型基金。
将贡献度按照从大到小的顺序排序后，步骤209具体包括：将贡献度排名靠前的预设个数的股票型基金，确定为符合预设表现稳定条件的目标股票型基金；或统计贡献度排名靠前的预定数量的目标基金发行公司；将预定数量的目标基金发行公司当前分别发行的股票型基金，确定为符合预设表现稳定条件的目标股票型基金。例如，若根据股票型基金排列顺序确定目标股票型基金，且预设提取的目标股票型基金数量为10，则按照贡献度排名，提取前10名股票型基金作为符合预设表现稳定条件的目标股票型基金。再例如，若根据基金发行公司排列顺序确定目标股票型基金，且预设提取的目标基金发行公司数量为10，则提取贡献度排名中前10名基金发行公司，将提取的基金发行公司当前分别发行的股票型基金，确定为符合预设表现稳定条件的目标股票型基金。
通过上述基金信息的分析方法，可以找出市场上收益表现具有稳定性、持久性和非随机性表现的目标股票型基金，并由其公开信息中的投资目标，可找出其股票投资策略，进而能筛选出目前在整个市场上相对有效的股票投资策略和表现稳定的股票型基金，并且能实时根据股票基金市场的变动得出不同的分析结果，使企业或投资者能最大程度的规避投资风险。
进一步的，为了提取出不同股票投资策略之间的相关性以及搭配效应，作为一种优选方式，本实施例还可包括：计算目标股票型基金与其他股票型基金之间的相关性信息；依据相关性信息，确定目标股票型基金的股票投资策略信息与其他股票型基金的股票投资策略信息之间的内在联系信息。其中，每只股票型基金的收益表现是与其策略一一对应，在计算股票型基金之间的协方差矩阵或者相关性矩阵，就可以通过收益之间的相关性，得到市场上各种基金策略之间的内在联系。相关性计算公式为：相关系数＝协方差/两个股票型基金的标准差之积。相关系数是度量两个随机变量间关联程度的量，取值范围为(-1，+1)，当相关系数小于0时，称为负相关；大于0时，称为正相关；等于0时，称为零相关。例如，在同一时间段内，基金A使用的跟踪创业板的被动投资策略，基金B使用的是跟踪上证50的被动投资策略，两只基金在这个时间段内的相关系数是-0.8，那么，可以得出一个结论是基金A的策略和基金B的策略是负相关的，对于基金中的基金(Fund of Funds，FOF)来说，可以同时持有基金A和B，从而实现风险对冲；如果分析出基金A和B为市场上表现稳定的基金，那么可以确定二者相关性计算结果相对可靠。
通过上述方法可进一步确定目标股票型基金的股票投资策略信息与其他股票型基金的股票投资策略信息之间的内在联系信息，从而得到更多的股票投资策略信息，分析出股票型基金之间的相关性，进而能提取出不同策略之间的市场表现及搭配效应，为企业或个人在股票型基金的选择提供更多可能。
进一步的，作为图1和图2所示方法的具体体现，本申请实施例提供了一种基金信息的分析装置，如图3所示，该装置包括：获取单元31、处理单元32、计算单元33、筛选单元34、分析单元35。
获取单元31，可用于获取当前流通的各只股票型基金对应的资讯数据，资讯数据中包含各只股票对应的每日收盘净值信息；
处理单元32，可用于按照各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息；
计算单元33，可用于计算对数收益率序列矩阵信息对应的样本协方差矩阵信息；
筛选单元34，可用于筛选样本协方差矩阵信息对应的偏离特征向量信息；
分析单元35，可用于分析各个股票型基金分别对偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金，以便获取目标股票型基金对应的股票投资策略信息。
其中，分析单元35，具体可用于将各个股票型基金按照基金发行公司进行归类，并配置每个基金发行公司对应的归类编码；按照第一预设公式
Figure PCTCN2019073288-appb-000016
计算各个股票型基金对偏离特征向量信息的贡献度，其中，
Figure PCTCN2019073288-appb-000017
为归类编码为s的股票型基金对偏离特征向量信息中排序第k大的偏离特征向量u (k)的贡献度；若指数型基金i的归类编码为s，则
Figure PCTCN2019073288-appb-000018
n s代表归类编码为s的股票型基金的基金数量，若指数型基金i的归类编码不为s，则
Figure PCTCN2019073288-appb-000019
Figure PCTCN2019073288-appb-000020
为u (k)的第i个分量；将贡献度按照从大到小的顺序排序，并依据贡献度排名靠前的预设个数的股票型基金，确定符合预设表现稳定条件的目标股票型基金。
在具体的应用场景中，分析单元35，具体还可用于将预设个数的股票型基金，确定为符合预设表现稳定条件的目标股票型基金；或统计贡献度排名靠前的预定数量的目标基金发行公司；将预定数量的目标基金发行公司当前分别发行的股票型基金，确定为符合预设表现稳定条件的目标股票型基金。
在具体的应用场景中，处理单元32具体可用于依据每日收盘净值信息中各只股票型基金的每日收盘净值为参数，按照第二预设公式
Figure PCTCN2019073288-appb-000021
计算对数收益率序列矩阵A；其中，
Figure PCTCN2019073288-appb-000022
Figure PCTCN2019073288-appb-000023
为股票i的对数收益率的标准差，N为股票数量。
相应的，计算单元33，具体可用于按照第三预设公式
Figure PCTCN2019073288-appb-000024
计算对数收益率序列矩阵A对应的样本协方差矩阵∑ N*N，A′为A的转置矩阵。
在具体的应用场景中，筛选单元34，具体可用于计算样本协方差矩阵∑ N*N的各个特征值以及各个特征值相应的特征向量；若样本协方差矩阵∑ N*N中的元素符合独立同分布，则利用第四预设公式
Figure PCTCN2019073288-appb-000025
计算样本协方差矩阵∑ N*N的特征值的理论最大边界值，λ max为理论最大边界值，Q由股票型基金的数量除以T得到；将各个特征值中大于λ max的特征值确定为异常特征值，并将异常特征值相应的特征向量确定为样本协方差矩阵信息对应的偏离特征向量。
在具体的应用场景中，为了更好的优化方案，进一步确定目标股票型基金的股票投资策略信息与其他股票型基金的股票投资策略信息之间的内在联系信息，如图4所示，本装置还可包括：确定单元36。
计算单元33，具体还可用于计算目标股票型基金与其他股票型基金之间的相关性信息；确定单元36，可用于依据相关性信息，确定目标股票型基金的股票投资策略信息与其他股票型基金的股票投资策略信息之间的内在联系信息。
需要说明的是，本实施例提供的一种基金信息的分析装置所涉及各功能单元的其他相应描述，可以参考图1至图2的对应描述，在此不再赘述。
基于上述如图1至图2所示方法，相应的，本实施例还提供了一种非易失性可读存储介质，其上存储有计算机可读指令，该可读指令被处理器执行时实现上述如图1至图2所示的基金信息的分析方法。
基于这样的理解，本申请的技术方案可以以软件产品的形式体现出来，该软件产品可以存储在一个非易失性存储介质(可以是CD-ROM，U盘，移动硬盘等)中，包括若干指令用以使得一台计算机设备(可以是个人计算机，服务器，或者网络设备等)执行本申请各个实施场景所述的方法。
基于上述如图1至图2所示的方法和图3、图4所示的虚拟装置实施例，为了实现上述目的，本实施例还提供了一种计算机设备，该计算机设备包括非易失性可读存储介质和处理器；非易失性可读存储介质，用于存储计算机可读指令；处理器，用于执行计算机可读指令以实现上述如图1至图2所示的 基金信息的分析方法。
可选的，该计算机设备还可以包括用户接口、网络接口、摄像头、射频(Radio Frequency，RF)电路，传感器、音频电路、WI-FI模块等等。用户接口可以包括显示屏(Display)、输入单元比如键盘(Keyboard)等，可选用户接口还可以包括USB接口、读卡器接口等。网络接口可选的可以包括标准的有线接口、无线接口(如WI-FI接口)等。
本领域技术人员可以理解，本实施例提供的一种计算机设备结构并不构成对该实体设备的限定，可以包括更多或更少的部件，或者组合某些部件，或者不同的部件布置。
非易失性存储介质中还可以包括操作系统、网络通信模块。操作系统是管理上述计算机设备硬件和软件资源的计算机可读指令，支持信息处理计算机可读指令以及其它软件和/或计算机可读指令的运行。网络通信模块用于实现非易失性存储介质内部各组件之间的通信，以及与信息处理实体设备中其它硬件和软件之间通信。
通过以上的实施方式的描述，本领域的技术人员可以清楚地了解到本申请可以借助软件加必要的通用硬件平台的方式来实现，也可以通过硬件实现。通过应用本申请的技术方案，与目前现有技术相比，本申请可通过获取当前流通的各只股票型基金对应的资讯数据，并利用咨询数据中包含的每日收盘净值信息得到对数收益率序列矩阵信息，依次计算出样本协方差矩阵信息，筛选出偏离特征向量信息，最后通过分析各个股票型基金分别对偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金，获取到目标股票型基金对应的股票投资策略信息，非常的方便且具有科学性，并且可以随时根据股票型基金的市场变化，实时得到对应的结果，为企业或个人对股票型基金的选择上提供重要的参考资料。并且还可以计算出目标股票型基金与其他股票型基金之间的相关性信息，确定目标股票型基金的股票投资策略信息与其他股票型基金的股票投资策略信息之间的内在联系信息，进而能够股票投资策略信息进行丰富扩展，提取出不同策略之间的市场表现及搭配效应。
本领域技术人员可以理解附图只是一个优选实施场景的示意图，附图中的模块或流程并不一定是实施本申请所必须的。本领域技术人员可以理解实施场景中的装置中的模块可以按照实施场景描述进行分布于实施场景的装置中，也可以进行相应变化位于不同于本实施场景的一个或多个装置中。上述实施场景的模块可以合并为一个模块，也可以进一步拆分成多个子模块。
上述本申请序号仅仅为了描述，不代表实施场景的优劣。以上公开的仅为本申请的几个具体实施场景，但是，本申请并非局限于此，任何本领域的技术人员能思之的变化都应落入本申请的保护范围。

Claims (20)

  1. 一种基金信息的分析方法，其特征在于，包括：
    获取当前流通的各只股票型基金对应的资讯数据，所述资讯数据中包含所述各只股票型基金对应的每日收盘净值信息；
    按照所述各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息；
    计算所述对数收益率序列矩阵信息对应的样本协方差矩阵信息；
    筛选所述样本协方差矩阵信息对应的偏离特征向量信息；
    分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金。
  2. 根据权利要求1的方法，其特征在于，所述分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金，具体包括：
    将所述各个股票型基金按照基金发行公司进行归类，并配置每个所述基金发行公司对应的归类编码；
    按照第一预设公式
    Figure PCTCN2019073288-appb-100001
    计算所述各个股票型基金对所述偏离特征向量信息的贡献度，其中，
    Figure PCTCN2019073288-appb-100002
    为归类编码为s的股票型基金对所述偏离特征向量信息中排序第k大的偏离特征向量u (k)的贡献度；若指数型基金i的归类编码为s，则
    Figure PCTCN2019073288-appb-100003
    n s代表归类编码为s的股票型基金的基金数量，若指数型基金i的归类编码不为s，则
    Figure PCTCN2019073288-appb-100004
    为u (k)的第i个分量；
    将所述贡献度按照从大到小的顺序排序，并依据所述贡献度排名靠前的预设个数的股票型基金，确定符合所述预设表现稳定条件的目标股票型基金。
  3. 根据权利要求2所述的方法，其特征在于，所述依据所述贡献度排名靠前的预设个数的股票型基金，确定符合所述预设表现稳定条件的目标股票型基金，具体包括：
    将所述预设个数的股票型基金，确定为符合所述预设表现稳定条件的目标股票型基金；或统计所述贡献度排名靠前的预定数量的目标基金发行公司；
    将所述预定数量的目标基金发行公司当前分别发行的股票型基金，确定为符合所述预设表现稳定条件的目标股票型基金。
  4. 根据权利要求1所述的方法，其特征在于，所述按照所述各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息，具体包括：
    依据所述每日收盘净值信息中所述各只股票型基金的每日收盘净值为参数，按照第二预设公式
    Figure PCTCN2019073288-appb-100005
    计算对数收益率序列矩阵A；
    其中，
    Figure PCTCN2019073288-appb-100006
    是股票i在第t日的收盘净值，
    Figure PCTCN2019073288-appb-100007
    是股票i在第t-1日的收盘净值，r i (t)是股票i在第t日的对数收益率，
    Figure PCTCN2019073288-appb-100008
    T为观测时点个数，
    Figure PCTCN2019073288-appb-100009
    δ i为股票i的对数收益率的标准差，N为股票数量。
  5. 根据权利要求4所述的方法，其特征在于，所述计算所述对数收益率序列矩阵信息对应的样本协方差矩阵信息，具体包括：
    按照第三预设公式
    Figure PCTCN2019073288-appb-100010
    计算所述对数收益率序列矩阵A对应的样本协方差矩阵∑ N*N，A′为A的转置矩阵。
  6. 根据权利要求5所述的方法，其特征在于，所述筛选所述样本协方差矩阵信息对应的偏离特征向量信息，具体包括：
    计算所述样本协方差矩阵∑ N*N的各个特征值以及所述各个特征值相应的特征向量；
    若所述样本协方差矩阵∑ N*N中的元素符合独立同分布，则利用第四预设公式
    Figure PCTCN2019073288-appb-100011
    计算所述样本协方差矩阵∑ N*N的特征值的理论最大边界值，所述λ max为所述理论最大边界值，Q由股票型基金的数量除以T得到；
    将所述各个特征值中大于所述λ max的特征值确定为异常特征值，并将所述异常特征值相应的特征向量确定为所述样本协方差矩阵信息对应的偏离特征向量。
  7. 根据权利要求1所述的方法，其特征在于，所述分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金之后，所述方法还包括：
    计算所述目标股票型基金与其他股票型基金之间的相关性信息；
    依据所述相关性信息，确定所述目标股票型基金的股票投资策略信息与所述其他股票型基金的股票投资策略信息之间的内在联系信息。
  8. 一种基金信息的分析装置，其特征在于，包括：
    获取单元，用于获取当前流通的各只股票型基金对应的资讯数据，所述资讯数据中包含所述各只股票对应的每日收盘净值信息；
    处理单元，用于按照所述各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息；
    计算单元，用于计算所述对数收益率序列矩阵信息对应的样本协方差矩阵信息；
    筛选单元，用于筛选所述样本协方差矩阵信息对应的偏离特征向量信息；
    分析单元，用于分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现 稳定条件的目标股票型基金。
  9. 根据权利要求8的装置，其特征在于，
    所述分析单元，具体用于将所述各个股票型基金按照基金发行公司进行归类，并配置每个所述基金发行公司对应的归类编码；
    按照第一预设公式
    Figure PCTCN2019073288-appb-100012
    计算所述各个股票型基金对所述偏离特征向量信息的贡献度，其中，
    Figure PCTCN2019073288-appb-100013
    为归类编码为s的股票型基金对所述偏离特征向量信息中排序第k大的偏离特征向量u (k)的贡献度；若指数型基金i的归类编码为s，则
    Figure PCTCN2019073288-appb-100014
    n s代表归类编码为s的股票型基金的基金数量，若指数型基金i的归类编码不为s，则
    Figure PCTCN2019073288-appb-100015
    为u (k)的第i个分量；
    将所述贡献度按照从大到小的顺序排序，并依据所述贡献度排名靠前的预设个数的股票型基金，确定符合所述预设表现稳定条件的目标股票型基金。
  10. 根据权利要求9所述的装置，其特征在于，
    所述分析单元，具体还用于将所述预设个数的股票型基金，确定为符合所述预设表现稳定条件的目标股票型基金；或
    统计所述贡献度排名靠前的预定数量的目标基金发行公司；
    将所述预定数量的目标基金发行公司当前分别发行的股票型基金，确定为符合所述预设表现稳定条件的目标股票型基金。
  11. 根据权利要求8所述的装置，其特征在于，
    所述处理单元，具体用于依据所述每日收盘净值信息中所述各只股票型基金的每日收盘净值为参数，按照第二预设公式
    Figure PCTCN2019073288-appb-100016
    计算对数收益率序列矩阵A；
    其中，
    Figure PCTCN2019073288-appb-100017
    是股票i在第t日的收盘净值，
    Figure PCTCN2019073288-appb-100018
    是股票i在第t-1日的收盘净值，r i (t)是股票i在第t日的对数收益率，
    Figure PCTCN2019073288-appb-100019
    T为观测时点个数，
    Figure PCTCN2019073288-appb-100020
    δ i为股票i的对数收益率的标准差，N为股票数量。
  12. 根据权利要求11所述的装置，其特征在于，
    所述计算单元，具体用于按照第三预设公式
    Figure PCTCN2019073288-appb-100021
    计算所述对数收益率序列矩阵A对应的样本协方差矩阵∑ N*N，A′为A的转置矩阵。
  13. 根据权利要求12所述的装置，其特征在于，
    所述筛选单元，具体用于计算所述样本协方差矩阵∑ N*N的各个特征值以及所述各个特征值相应的特征向量；
    若所述样本协方差矩阵∑ N*N中的元素符合独立同分布，则利用第四预设公式
    Figure PCTCN2019073288-appb-100022
    计算所述样本协方差矩阵∑ N*N的特征值的理论最大边界值，所述λ max为所述理论最大边界值，Q由股票型基金的数量除以T得到；
    将所述各个特征值中大于所述λ max的特征值确定为异常特征值，并将所述异常特征值相应的特征向量确定为所述样本协方差矩阵信息对应的偏离特征向量。
  14. 根据权利要求8所述的装置，其特征在于，所述装置还包括：确定单元；
    所述计算单元，还用于计算所述目标股票型基金与其他股票型基金之间的相关性信息；
    所述确定单元，用于依据所述相关性信息，确定所述目标股票型基金的股票投资策略信息与所述其他股票型基金的股票投资策略信息之间的内在联系信息。
  15. 一种非易失性可读存储介质，其上存储有计算机可读指令，其特征在于，所述计算机可读指令被处理器执行时实现基金信息的分析方法，包括：
    获取当前流通的各只股票型基金对应的资讯数据，所述资讯数据中包含所述各只股票型基金对应的每日收盘净值信息；按照所述各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息；计算所述对数收益率序列矩阵信息对应的样本协方差矩阵信息；筛选所述样本协方差矩阵信息对应的偏离特征向量信息；分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金。
  16. 根据权利要求15的非易失性可读存储介质，其特征在于，所述计算机可读指令被处理器执行时实现所述分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金，具体包括：
    将所述各个股票型基金按照基金发行公司进行归类，并配置每个所述基金发行公司对应的归类编码；按照第一预设公式
    Figure PCTCN2019073288-appb-100023
    计算所述各个股票型基金对所述偏离特征向量信息的贡献度，其中，
    Figure PCTCN2019073288-appb-100024
    为归类编码为s的股票型基金对所述偏离特征向量信息中排序第k大的偏离特征向量u (k)的贡献度；若指数型基金i的归类编码为s，则
    Figure PCTCN2019073288-appb-100025
    n s代表归类编码为s的股票型基金的基金数量，若指数型基金i的归类编码不为s，则
    Figure PCTCN2019073288-appb-100026
    为u (k)的第i个分量；将所述贡献度按照从大到小的顺序排序，并依据所述贡献度排名靠前的预设个数的股票型基金，确定符合所述预设表现稳定条件的目标股票型基金。
  17. 根据权利要求16所述的非易失性可读存储介质，其特征在于，所述计算机可读指令被处理器执行时实现所述依据所述贡献度排名靠前的预设个数的股票型基金，确定符合所述预设表现稳定条件的目标股票型基金，具体包括：
    将所述预设个数的股票型基金，确定为符合所述预设表现稳定条件的目标股票型基金；或统计所 述贡献度排名靠前的预定数量的目标基金发行公司；将所述预定数量的目标基金发行公司当前分别发行的股票型基金，确定为符合所述预设表现稳定条件的目标股票型基金。
  18. 一种计算机设备，包括非易失性可读存储介质、处理器及存储在非易失性可读存储介质上并可在处理器上运行的计算机可读指令，其特征在于，所述处理器执行所述计算机可读指令时实现基金信息的分析方法，包括：
    获取当前流通的各只股票型基金对应的资讯数据，所述资讯数据中包含所述各只股票型基金对应的每日收盘净值信息；按照所述各只股票型基金对应的每日收盘净值信息进行标准化处理，得到对数收益率序列矩阵信息；计算所述对数收益率序列矩阵信息对应的样本协方差矩阵信息；筛选所述样本协方差矩阵信息对应的偏离特征向量信息；分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金。
  19. 根据权利要求18的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时实现所述分析各个股票型基金分别对所述偏离特征向量信息的贡献度，得到符合预设表现稳定条件的目标股票型基金，具体包括：
    将所述各个股票型基金按照基金发行公司进行归类，并配置每个所述基金发行公司对应的归类编码；按照第一预设公式
    Figure PCTCN2019073288-appb-100027
    计算所述各个股票型基金对所述偏离特征向量信息的贡献度，其中，
    Figure PCTCN2019073288-appb-100028
    为归类编码为s的股票型基金对所述偏离特征向量信息中排序第k大的偏离特征向量u (k)的贡献度；若指数型基金i的归类编码为s，则
    Figure PCTCN2019073288-appb-100029
    n s代表归类编码为s的股票型基金的基金数量，若指数型基金i的归类编码不为s，则
    Figure PCTCN2019073288-appb-100030
    为u (k)的第i个分量；将所述贡献度按照从大到小的顺序排序，并依据所述贡献度排名靠前的预设个数的股票型基金，确定符合所述预设表现稳定条件的目标股票型基金。
  20. 根据权利要求19所述的计算机设备，其特征在于，所述处理器执行所述计算机可读指令时实现所述依据所述贡献度排名靠前的预设个数的股票型基金，确定符合所述预设表现稳定条件的目标股票型基金，具体包括：
    将所述预设个数的股票型基金，确定为符合所述预设表现稳定条件的目标股票型基金；或统计所述贡献度排名靠前的预定数量的目标基金发行公司；将所述预定数量的目标基金发行公司当前分别发行的股票型基金，确定为符合所述预设表现稳定条件的目标股票型基金。
PCT/CN2019/073288 2018-10-30 2019-01-27 基金信息的分析方法及装置、存储介质、计算机设备 Ceased WO2020087791A1 (zh)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
CN201811280448.8 2018-10-30
CN201811280448.8A CN109636615A (zh) 2018-10-30 2018-10-30 基金信息的分析方法及装置、存储介质、计算机设备

Publications (1)

Publication Number Publication Date
WO2020087791A1 true WO2020087791A1 (zh) 2020-05-07

Family

ID=66066896

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/CN2019/073288 Ceased WO2020087791A1 (zh) 2018-10-30 2019-01-27 基金信息的分析方法及装置、存储介质、计算机设备

Country Status (2)

Country Link
CN (1) CN109636615A (zh)
WO (1) WO2020087791A1 (zh)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103455943A (zh) * 2013-09-02 2013-12-18 深圳市国泰安信息技术有限公司 一种股票或股票投资组合波动率的预测方法、装置
CN105989535A (zh) * 2015-02-10 2016-10-05 上海华颂软件科技有限公司 一种股票波动率预测方法及系统
CN106934503A (zh) * 2017-03-22 2017-07-07 四川倍发科技有限公司 一种股票波动率预测系统
CN108090834A (zh) * 2017-12-18 2018-05-29 孙嘉 基金评价方法和装置

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
CN103455943A (zh) * 2013-09-02 2013-12-18 深圳市国泰安信息技术有限公司 一种股票或股票投资组合波动率的预测方法、装置
CN105989535A (zh) * 2015-02-10 2016-10-05 上海华颂软件科技有限公司 一种股票波动率预测方法及系统
CN106934503A (zh) * 2017-03-22 2017-07-07 四川倍发科技有限公司 一种股票波动率预测系统
CN108090834A (zh) * 2017-12-18 2018-05-29 孙嘉 基金评价方法和装置

Also Published As

Publication number Publication date
CN109636615A (zh) 2019-04-16

Similar Documents

Publication Publication Date Title
CN107679946B (zh) 基金产品推荐方法、装置、终端设备及存储介质
Huber et al. Radius matching on the propensity score with bias adjustment: tuning parameters and finite sample behaviour
WO2019214248A1 (zh) 一种风险评估方法、装置、终端设备及存储介质
Zekić-Sušac et al. Predicting company growth using logistic regression and neural networks
CN113516511B (zh) 一种金融产品购买预测方法、装置及电子设备
Acheampong et al. Foreign direct investment (FDI) inflows into Ghana: Should the focus be on infrastructure or natural resources? Short-run and long-run analyses
Gagliardini et al. Comparing asset pricing models by the conditional Hansen-Jagannathan distance
CN110737917A (zh) 基于隐私保护的数据共享装置、方法及可读存储介质
Mumtaz et al. The determinants of Chinese outward foreign direct investment: A closer look
Klein et al. Investment in non-fungible tokens (NFTs): the return of Ethereum secondary market NFT sales: NC Klein et al.
CN109978300B (zh) 客户风险承受力量化方法及系统、资产配置方法及系统
Ghosh et al. Causes and consequences of the CEO also being the chair of the board
Yuan A new model for the joint valuation of S&P 500 and VIX options: Specification analysis
Bahadur et al. Dimenslon estlmatlon of equlty markets
Muvingi et al. Default prediction models a comparison between market based models and accounting based: Case of the Zimbabwe Stock Exchange 2010-2013
CN113222767A (zh) 指数化证券组合的数据处理方法及装置
WO2020087791A1 (zh) 基金信息的分析方法及装置、存储介质、计算机设备
Wesselhöfft et al. Risk-Constrained Kelly portfolios under alpha-stable laws
Wimmer et al. Applying information technology to financial statement analysis for market capitalization prediction
CN114154590A (zh) 用户聚类方法、装置、终端设备以及存储介质
Fletcher Examination of conditional asset pricing in UK stock returns
Bao et al. Sparse Interval-valued Time Series Modeling with Machine Learning
Motoc The Influence of Audit Opinion on Companies’ Insolvency Risk. Evidence for Romania
Jay et al. Robust covariance matrix estimation and portfolio allocation: the case of non-homogeneous assets
Keranović et al. Estimating the Number of Latent Factors in High-Dimensional Financial Time Series

Legal Events

Date Code Title Description
121 Ep: the epo has been informed by wipo that ep was designated in this application

Ref document number: 19878297

Country of ref document: EP

Kind code of ref document: A1

NENP Non-entry into the national phase

Ref country code: DE

122 Ep: pct application non-entry in european phase

Ref document number: 19878297

Country of ref document: EP

Kind code of ref document: A1