WO2021060968A1 - Système et procédé permettant de fournir une recommandation de marché boursier - Google Patents

Système et procédé permettant de fournir une recommandation de marché boursier Download PDF

Info

Publication number
WO2021060968A1
WO2021060968A1 PCT/MY2020/050059 MY2020050059W WO2021060968A1 WO 2021060968 A1 WO2021060968 A1 WO 2021060968A1 MY 2020050059 W MY2020050059 W MY 2020050059W WO 2021060968 A1 WO2021060968 A1 WO 2021060968A1
Authority
WO
WIPO (PCT)
Prior art keywords
stock
recommendation
analyst
published
accuracy
Prior art date
Application number
PCT/MY2020/050059
Other languages
English (en)
Inventor
Yasaman EFTEKHARYPOUR
Weiying KOK
Chuan Hai NGO
Original Assignee
Mimos Berhad
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 Mimos Berhad filed Critical Mimos Berhad
Publication of WO2021060968A1 publication Critical patent/WO2021060968A1/fr

Links

Classifications

    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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
    • G06Q30/00—Commerce
    • G06Q30/06—Buying, selling or leasing transactions
    • G06Q30/0601—Electronic shopping [e-shopping]
    • G06Q30/0631—Item recommendations
    • G—PHYSICS
    • G06—COMPUTING; CALCULATING OR 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

Definitions

  • the present invention relates generally to stock market recommendation. More particularly, the present invention relates to an improved system and method for providing a stock market recommendation using a comprehensive behavior prediction approach.
  • stock market or interchangeably known as stock exchange refers to the collection of markets and exchanges where regular activities of buying, selling, and issuance of shares of publicly-held companies take place. Such financial activities are conducted through institutionalized formal exchanges or over-the- counter marketplaces which operate under a defined set of regulations. There can be multiple stock trading venues in a country or a region which allow transactions in stocks and other forms of securities.
  • Portfolio managers are professional who invest portfolios, or collections of securities for clients. They obtain stock recommendations from stock analysts and make the buy or sell decisions for the portfolio. The stock analysts are usually employed to research publicly-traded companies and attempt to forecast whether a company’s stock is likely to rise or fall in price.
  • the prediction of future behavior of stock market is based only on historical numeric values of market entities alone.
  • the problem with this approach is that it fails to capture immediate and unexpected changes that are possibly reflected in many media. Relying on media sentiments arising from media such as social media and news sources with various negative and positive reviews, may however render the entire stock market or behavior prediction to become hampered.
  • the prediction may be, unintentionally, stuck in N atmosphere that is managed (and manipulated) by one or more groups to deliberately publish content for a particular personal goal.
  • United States Patent No. 8,285,619 B2 discloses a method of using natural language processing (NLP) techniques to extract information from online news feeds and then using the extracted information to predict changes in stock prices or volatilities. According to the ‘619 patent, these predictions can be used to make profitable trading strategies. It was disclosed in the ‘619 patent that company names can be recognized and simple templates describing company actions can be automatically filled using parsing or pattern matching on words in or near the sentence containing the company name. These templates can be clustered into groups which are statistically correlated with changes in the stock prices.
  • NLP natural language processing
  • the system is composed with two parts: message understanding component that automatically fills in simple templates and a statistical correlation component that tests the correlation of these patterns to increase or decrease in the stock price.
  • message understanding component that automatically fills in simple templates
  • statistical correlation component that tests the correlation of these patterns to increase or decrease in the stock price.
  • the methods can be applied to a broad range of text, including articles in online newspapers such as the Wall Street Journal, financial newsletters, radio &TV transcripts and annual reports.
  • statistical patterns in Internet usage data and Internet data such as newly released textual information on Web pages are further leveraged.
  • the present invention provides a system for providing a stock market recommendation.
  • the system of the present invention may be characterized by an analyst recommendation unit for providing a first stock recommendation of a stock at a selected day, a stock behavior time series model unit for determining a predicted stock price movement or behavior of the stock for the selected day using a time series model trained on historical stock prices, a media-based behavior predictor unit for determining a media mood score of the stock using a media mood time series model trained on published media at the selected day and an aggregator unit for providing a second stock recommendation of the stock by way of aggregating the first stock recommendation, the predicted stock price movement or behavior and the media mood score using a predefined statistical model.
  • the first stock recommendation is derived from data selected from a group comprising published analyst recommendations, cross-correlation graphs and partial correlation networks established for the stock thereof.
  • the first stock recommendation is selected based on an accuracy value computed for each stock analyst associated with the published analyst recommendations, a primary alternative stock recommendation of a primary alternative stock identified from the cross-correlation graphs, or a secondary alternative stock recommendation of a secondary alternative stock identified from the partial correlation networks.
  • the analyst recommendation unit comprises an analyst recommendation mapper unit, an accuracy table unit connected to the analyst mapper recommendation unit, a cross-correlation graph unit and a partial correlation unit.
  • the analyst recommendation mapper unit is configured for mapping the published analyst recommendations issued by each stock analyst on a calendar.
  • the accuracy table unit is configured to determine the accuracy value for each stock analyst by way of establishing an accuracy function using accuracy measure based on a true-false test; conducting the true-false test by comparing the published analyst recommendations against actual stock movements or behaviors for the stock of a same day, if the published analyst recommendations match actual stock movements or behaviors, then assigns a true value; and if the published analyst recommendations do not match actual stock movements or behaviors, then assigns a false value; determining a number of the true value and the false value generated from the true-false test thereby entering the same to the accuracy function; and determining the accuracy value of each stock analyst.
  • the accuracy table unit identifies a stock analyst with the highest accuracy value and assists in providing the first stock recommendation of the stock at the selected day by retrieving the stock analyst’s published analyst recommendations for the stock thereof.
  • the cross-correlation graph unit is configured to provide the cross correlation graphs of stocks comprising cross-correlation values between the stocks therein adapted to identify the primary alternative stock thereof.
  • the partial correlation unit is configured for providing the partial correlation networks comprising partial correlation values adapted to identify the secondary alternative stock being the most influential stock on the stock thereof and the primary alternative stock thereof.
  • the present invention provides a method of providing a stock market recommendation.
  • the method may be characterized by the steps of providing a first stock recommendation of a stock at a selected day, comprising deriving the first stock recommendation from data selected from a group comprising published analyst recommendations, cross-correlation graphs and partial correlation networks established for the stock thereof, including selecting the first stock recommendation based on an accuracy value computed for each stock analyst associated with the published analyst recommendations, or selecting the first stock recommendation based on a primary alternative stock recommendation of a primary alternative stock identified from the cross-correlation graphs, or selecting the first stock recommendation based on a secondary alternative stock recommendation of a secondary alternative stock identified from the partial correlation networks; determining a predicted stock price movement or behavior of the stock for the selected day using a time series model trained on historical stock prices; determining a media mood score of the stock using a media mood time series model trained on published media at the selected day; and providing a second stock recommendation of the stock by way of aggregating the first stock recommendation, the predicted stock price movement or behavior and the media mood score using a predefined statistical model.
  • the step of selecting the first stock recommendation based on an accuracy value comprising establishing an accuracy function using accuracy measure based on a true-false test; conducting the true-false test by comparing the published analyst recommendations against actual stock movements or behaviors for the stock of a same day, if the published analyst recommendations match actual stock movements or behaviors, then assigns a true value; and if the published analyst recommendations do not match actual stock movements or behaviors, then assigns a false value; determining a number of the true value and the false value generated from the true-false test thereby entering the same to the accuracy function; and determining the accuracy value of each stock analyst.
  • the step of selecting the first stock recommendation based on an accuracy value further comprising identifying a stock analyst with the highest accuracy value; and assisting to provide the first stock recommendation of the stock at the selected day by retrieving the stock analyst’s published analyst recommendations for the stock thereof.
  • the step of selecting the first stock recommendation based on a primary alternative stock recommendation comprising providing the cross correlation graphs of stocks comprising cross-correlation values between the stocks therein; and identifying the primary alternative stock based on the cross-correlation values thereof.
  • the step of selecting the first stock recommendation based on a secondary alternative stock recommendation comprising providing the partial correlation networks of stocks comprising partial correlation values; and identifying the secondary alternative stock being the most influential stock on the stock thereof and the primary alternative stock thereof. It is an objective of the present invention to improve the stock recommendation by way of predicting stock market behavior using a comprehensive behavior prediction approach that incorporates correlation between any two market entities, partial correlation value capable of detecting the most influential stock on a pair of stock relationship, information extractable from media and historical published prediction report from expert analyst.
  • Figure 1 shows an arrangement of a system for providing a stock recommendation according to one embodiment of the present invention
  • Figure 2 is a flow diagram of a method comprising steps providing a stock recommendation according to one embodiment of the present invention
  • Figure 3 is a flow diagram of the step of providing a first stock recommendation of a stock at a selected day according to one embodiment of the present invention
  • Figure 4 is an example of correlation matrix represented as graph according to one embodiment of the present invention.
  • Figure 5 is an example of partial correlation network according to one embodiment of the present invention
  • Figure 6 illustrates the step of mapping the published analyst recommendations issued by each stock analyst on a calendar by an analyst recommendation mapper unit according to one embodiment of the present invention
  • Figures 7 and 8 illustrate the steps of selecting the first stock recommendation based on an accuracy value and determining an accuracy value for each stock analyst by an accuracy table unit according to one embodiment of the present invention
  • Figure 9 is a flow diagram depicting the step of finding pairs of stocks that are correlated with cross-correlation values either positive or negative by a cross correlation graph unit according to one embodiment of the present invention.
  • Figure 10 is a flow diagram depicting the step of finding the most influential stock by a partial correlation unit according to one embodiment of the present invention.
  • Figure 11 is a flow diagram depicting the step of updating an accuracy table by an accuracy table unit according to one embodiment of the present invention.
  • Figure 12 is a flow diagram depicting the step of retraining each analyst model according to one embodiment of the present invention.
  • Figure 13 is a flow diagram depicting the step of selecting the first stock recommendation based on a primary alternative stock recommendation according to another embodiment of the present invention.
  • Figure 14 is a flow diagram depicting the step of selecting the first stock recommendation based on the most accurate analyst model recorded for the stock thereof according to another embodiment of the present invention.
  • Figure 15 is a flow diagram depicting the step of selecting the first stock recommendation based on a secondary alternative stock recommendation according to another embodiment of the present invention.
  • the present invention employs a comprehensive behavior prediction approach using a correlation between any two market entities, a partial correlation value that is capable of detecting the most influential stock on a pair of stock relationship, an information extractable from media and a historical published prediction report from an expert analyst. It is noted that based on the above comprehensive behavior prediction approach the present invention is capable of capturing immediate and unexpected changes reflected in media in addition to the historical numeric values of market entities.
  • the present invention provides a system for providing a stock market recommendation using the above-mentioned comprehensive behavior prediction approach.
  • the system of the present invention preferably comprises an analyst recommendation unit 100, a stock behavior time series model unit 200, a media-based behavior predictor unit 300 and an aggregator unit 400. It is preferred that the analyst recommendation unit 100, the stock behavior time series model unit 200, the media-based behavior predictor unit 300 and the aggregator unit 400 are located in a stock decisions recommender module.
  • the analyst recommendation unit 100 is preferably connected to the aggregator unit 400 which is also connected to the stock behavior time series model unit 200 and the media-based behavior predictor unit 300 thereof.
  • An arrangement of the system of the present invention is shown in Figure 1.
  • Figures 2 and 3 of the accompanying drawings provide flow diagrams depicting a method of providing the stock market recommendation that comprises steps executable by the system of the present invention.
  • the system of the present invention also comprises a processor, a memory unit and a storage unit connected to the processor thereof.
  • the processor, the memory unit and the storage unit are connected to the stock decisions recommender module through the processor thereof.
  • the memory unit may be configured as a temporary storage of graph data associated with at least one graph database during creation of at least one tree.
  • the at least one tree preferably has a plurality of associated memory cells containing values.
  • the processor may be configured to perform a plurality of calculations related to the graph data thereof based on a plurality of processor executable instructions.
  • the storage unit may be configured to store a plurality of serialized trees in form of a plurality of files to be used while performing an accelerated query execution specifying that queries are accelerated if they are eligible for acceleration.
  • the analyst recommendation unit 100 comprises an analyst recommendation mapper unit 101, an accuracy table unit 102, a cross-correlation graph unit 103 and a partial correlation unit 104.
  • the analyst recommendation unit 100 is configured for providing a first stock recommendation (also known as “initial stock recommendation”) of the stock at a selected day, i.e. present day (see step 500). It is preferred that the first stock recommendation is derived from data selected from a group comprising published analyst recommendations, cross-correlation graphs and partial correlation networks established for the stock thereof (see step 501).
  • the first stock recommendation is preferably selected based on an accuracy value that is computed for each stock analyst associated with the published analyst recommendations (see step 501a).
  • the first stock recommendation is preferably selected based on a primary alternative stock recommendation of a primary alternative stock identified from the cross-correlation graphs (see step 501 b).
  • Figure 4 exemplarily provides a cross-correlation matrix that is represented in a graph, i.e. cross-correlation graph.
  • the cross-correlation between two market entities is denoted by dotted lines and market entities like stocks and securities are denoted by a circle.
  • Stock 2 and Stock 1 in Figure 4 have a cross-correlation value of 0.1 in respect of each other.
  • the first stock recommendation is preferably selected based on a secondary alternative stock recommendation of a secondary alternative stock identified from the partial correlation networks (see step 501c).
  • Figure 5 exemplarily provides a partial correlation network of various stocks which is a statistical measure adapted to determine how the correlation between two variables is affected by a third variable. According to Figure 5, it can be seen that FRONTKN.KL and MUIIND-LA.KL are two stocks located in the center and probably have maximum influence on other stocks in the neighborhood.
  • the analyst recommendation mapper unit 101 of the analyst recommendation unit 100 may be configured for mapping the published analyst recommendations issued by each stock analyst on a calendar.
  • the calendar used herein may provide information such as the published analyst recommendations in timeline, calendar or other view for both past, ongoing and future tasks.
  • Figure 6 of the accompanying drawings illustrates the step of mapping the published analyst recommendations issued by each stock analyst onto the calendar by the analyst recommendation mapper unit 101.
  • three analyst recommendations published by three different stock analysts i.e. Analyst 16, Analyst 12 and Analyst 7, are mapped to the calendar based on their respective time frames.
  • Each analyst recommendation made by the stock analyst has a valid effective time frame, for example until 3-4 weeks, months, not more than 12 months or more than 12 months.
  • the accuracy table unit 102 of the analyst recommendation unit 100 may be configured for determining the accuracy value for each stock analyst as outlined in step 501a where the first stock recommendation is selected based on the accuracy value computed for each stock analyst associated with the published analyst recommendations.
  • the accuracy table unit 102 preferably establishes an accuracy function using accuracy measure based on a true-false test.
  • the true- false test will subsequently be conducted by way of comparing the published analyst recommendations against actual stock movements or behaviors for the stock in question of a same day.
  • the actual stock movements or behaviors may refer to the stock prices of the day, which can change (e.g. go up and down or remain) on a daily basis by market forces.
  • the accuracy table unit 102 assigns a true value. If otherwise, i.e. the published analyst recommendations do not match actual stock movements or behaviors, then a false value will be assigned.
  • the accuracy table unit 102 Upon completion of the true-false test, the accuracy table unit 102 will count and determine a number of the true value and the false value generated from the true-false test thereby entering the same to the accuracy function.
  • the accuracy function is provided as follows:
  • the accuracy table unit 102 will compute and determine the accuracy value of each stock analyst.
  • Figures 7 and 8 respectively illustrate the steps of selecting the first stock recommendation based on the accuracy value and determining an accuracy value for each stock analyst (see step 501a) by the accuracy table unit 102.
  • each true value and each false value carry a value with magnitude 1 and the number of the true value and the false value resulted from the true-false test will be determined for a particular period by computing a sum of those true value(s) and false value(s) accordingly.
  • the accuracy value for Analyst 2 for the same stock and period is computed at 0.6.
  • the stock analysts thereof i.e. Analyst 16, Analyst 12 and Analyst 7 will be subject to the true-false test alongside the accuracy function for determining the respective accuracy value.
  • the accuracy values of the stock analysts are crucial for the purpose of selecting a stock recommendation being the first stock recommendation for the stock at the selected day.
  • the present invention adopts a result lookup table such as Table 1 shown in Figure 7 to assist with the true-false test that runs by the accuracy table unit 102. Based on the true-false test conducted for each stock analyst, an analyst accuracy table for Stock 1 will be tabulated and completed as shown in Figure 7.
  • the analyst accuracy table lists down all the stock analysts and their prediction results from the true-false test which are ordered in accordance to number of days, weeks and months after the publication of report containing analyst recommendations. Likewise, all records for all stocks (e.g. Stock 2, Stock 3 and Stock 4) are kept accordingly.
  • the analyst accuracy table of each stock is preferably updated on a daily basis (see Figure 8) by way of comparison between the published analyst recommendations and the actual stock movements or behaviors for the stock of a same day. For every newly published analyst recommendation, each of analyst model will be re-trained using new information (e.g. financial data, recommendation action) extracted from the reports collected from online resources or any storage media. The new analyst model is employed to replace the old analyst model at the end of step of retraining each analyst model outlined in Figure 12.
  • new information e.g. financial data, recommendation action
  • an overall accuracy table is developed based on the analyst accuracy tables of the stocks thereof. For each particular period of time, a stock analyst with the highest accuracy value (as determined by the above-mentioned accuracy function) recorded for a particular stock will be identified and chosen by the accuracy table unit 102 as the most accurate stock analyst for that particular stock in that particular period of time. Subsequently, the accuracy table unit 102 assists in providing the selected stock analyst’s published analyst recommendation for the stock thereof. The published analyst recommendation of the said selected stock analyst of the highest accuracy value will be elected as the first stock recommendation for the stock at the selected day thereof by the accuracy table unit 102. Even though the overall accuracy table shown in Figure 8 is in continuity from Figure 7, the data (e.g. published analyst recommendations) adopted for calculating the accuracy value and henceforth the selection of the stock analyst with the highest accuracy value may differ as the data can be obtained from different sources, and not necessary reflecting the same data.
  • the cross-correlation unit 103 of the analyst recommendation unit 100 may be configured for providing the cross-correlation graphs of various stocks comprising cross-correlation values between the stocks therein.
  • the cross-correlation graphs are a statistical measure that determines how assets move in relation to each other.
  • a cross-correlation between stocks is measured on a scale of -1 to +1.
  • a perfect positive correlation between two stocks has a reading of +1.
  • a perfect negative correlation has a reading of -1.
  • Each cross-correlation graph developed at the end of this step will be stored in a suitable storage medium like the storage unit thereof.
  • the cross-correlation graph or the latest moving relation graph is preferably created based on a specific time window which can be determined based on the previously obtained optimized value or can be dynamically updated. It is an objective of the cross-correlation unit 103 to provide the cross-correlation graphs for identification of the primary alternative stock thereof as outlined in step 501b where the first stock recommendation is selected based on the primary alternative stock recommendation of the primary alternative stock that is identified from the cross-correlation graphs.
  • the cross-correlation unit 103 preferably provides the cross-correlation graphs of stocks comprising the cross-correlation values between the stocks therein as exemplarily shown in Figure 4. Based from the cross-correlation graphs and considering the cross-correlation values between the stocks, the cross-correlation unit 103 identifies the primary alternative stock, in particular the one with the highest cross-correlation value with respect to the stock in question.
  • the partial correlation unit 104 of the analyst recommendation unit 100 may be configured for providing the partial correlation networks or graphs comprising partial correlation values.
  • the partial correlation values may be adapted to identify the secondary alternative stock being the most influential stock on the stock thereof and the primary alternative stock thereof as outlined in step 501 c where the first stock recommendation is selected based on the secondary alternative stock recommendation of the secondary alternative stock identified from the partial correlation networks.
  • the partial correlation unit 104 preferably provides the partial correlation networks of stocks comprising the partial correlation values. Based from the partial correlation networks and considering the partial correlation values that indicate how the correlation between two stocks is affected by a third stock, the partial correlation unit 104 identifies the secondary alternative stock being the most influential stock on the stock in question and the primary alternative stock thereof.
  • Figure 10 shows the step of finding the most influential stock by the partial correlation unit 104.
  • Each partial correlation network developed at the end of this step will be stored in a suitable storage medium like the storage unit thereof.
  • the partial correlation network is preferably created based on a specific time window which can be determined based on the previously obtained optimized value or can be dynamically updated.
  • the stock behavior time series model unit 200 may be configured for determining a predicted stock price movement or behavior of the stock in question for the selected day as outlined in step 600.
  • the predicted stock price movement or behavior is preferably determined using a time series model that is trained on historical stock prices.
  • the time series model is preferably built using trends of the past stock prices to determine the future stock prices.
  • the media-based behavior predictor unit 300 may be configure to determine a media mood score of the stock in question as outlined in step 700.
  • the media mood score can be computed using a media mood time series model that is trained on published media at the selected day.
  • the media mood score is indicative of value of different mood dimensions based on the published media with regards to the stock in question at the selected day.
  • the aggregator unit 400 aggregates the first or initial stock recommendation, the predicted stock price movement or behavior and the media mood score and subjects them to a predefined statistical model for providing a second stock recommendation as outlined in step 800.
  • the second stock recommendation is preferably adapted as the stock market recommendation for the stock in question at the selected day thereof.
  • Figure 13 depicts the step of selecting the first stock recommendation based on the primary alternative stock recommendation of the primary alternative stock in the event that no analyst recommendation is found for the stock in question at the selected day.
  • the cross-correlation graph unit 103 of the analyst recommendation unit 100 is configured to find Stock B which is a stock with the highest correlation value with Stock A if there is no analyst recommendation found for Stock A. Subsequently, a stock recommendation for Stock B will be determined through its selected analyst using the analyst accuracy table thereof.
  • the stock recommendation for Stock B mentioned herein is essentially or may refer to the primary alternative stock recommendation.
  • the cross correlation graph unit 103 determines a stock recommendation for Stock A using the primary alternative stock recommendation thereof based on the corresponding cross correlation value that Stock B has with Stock A.
  • the stock recommendation for Stock A mentioned herein is essentially or may refer to the first stock recommendation.
  • Figure 14 depicts the step of selecting the first stock recommendation based on the most accurate analyst model recorded for the stock thereof in the event that no Stock B (i.e. the primary alternative stock) is found for that particular time frame.
  • the cross-correlation graph unit 103 is configured to find the most accurate analyst model recorded for Stock A (i.e. the stock in question at the selected day) if there is no Stock B correlated with Stock A found.
  • the most accurate analyst model found thereof will be used to provide a stock recommendation for Stock A by way importing information pertaining to Stock A into the most accurate analyst model thereof.
  • the stock recommendation for Stock A mentioned herein is essentially or may refer to the first stock recommendation.
  • Figure 15 depicts the step of selecting the first stock recommendation based on the secondary alternative stock recommendation of the secondary alternative stock in the event that the most accurate stock analyst for that particular stock in that particular period of time (supposedly determined by the accuracy table unit 102 of the analyst recommendation unit 100) is not found and that the primary alternative stock highly correlated with the stock in question (supposedly determined by the cross-correlation graph unit 103 of the analyst recommendation unit 100) is not found.
  • the partial correlation unit 104 is configured to find a third part stock known as “Stock C” (i.e. the secondary alternative stock) using the partial correlation networks thereof.
  • Stock C is essentially effective on correlation between Stock A and any other stocks.
  • Stock C with the secondary alternative stock recommendation as well as its accompanying analyst model shall be adapted as the secondary alternative stock for the purpose of determining the first stock recommendation for the stock in question, i.e. Stock A.
  • the partial correlation unit 104 determines a stock recommendation for Stock A by way importing information pertaining to Stock A into the analyst model thereof.
  • the stock recommendation for Stock A mentioned herein is essentially or may refer to the first stock recommendation.
  • inventive subject matter has been described with reference to specific example embodiments, various modifications and changes may be made to these embodiments without departing from the broader scope of embodiments of the present disclosure.
  • inventive subject matter may be referred to herein, individually or collectively, by the term “invention” merely for convenience and without intending to voluntarily limit the scope of this application to any single disclosure or inventive concept if more than one is, in fact, disclosed.
  • the term “or” may be construed in either an inclusive or exclusive sense. Moreover, plural instances may be provided for resources, operations, or structures described herein as a single instance. Additionally, boundaries between various resources, operations, modules, engines, and data stores are somewhat arbitrary, and particular operations are illustrated in a context of specific illustrative configurations. Other allocations of functionality are envisioned and may fall within a scope of various embodiments of the present disclosure. In general, structures and functionality presented as separate resources in the example configurations may be implemented as a combined structure or resource. Similarly, structures and functionality presented as a single resource may be implemented as separate resources. These and other variations, modifications, additions, and improvements fall within a scope of embodiments of the present disclosure as represented by the appended claims. The specification and drawings are, accordingly, to be regarded in an illustrative rather than a restrictive sense.
  • first means “first,” “second,” and so forth may be used herein to describe various elements, these elements should not be limited by these terms. These terms are only used to distinguish one element from another. For example, a first contact could be termed a second contact, and, similarly, a second contact could be termed a first contact, without departing from the scope of the present example embodiments. The first contact and the second contact are both contacts, but they are not the same contact.
  • the term “if’ may be construed to mean “when” or “upon” or “in response to determining” or “in response to detecting,” depending on the context.
  • the phrase “if it is determined” or “if [a stated condition or event] is detected” may be construed to mean “upon determining” or “in response to determining” or “upon detecting [the stated condition or event]” or “in response to detecting [the stated condition or event],” depending on the context.

Landscapes

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

Abstract

La présente invention concerne un système et un procédé pour fournir une recommandation de marché boursier à l'aide d'une approche de prédiction de comportement global. Le système comprend une unité de recommandation d'analyste pour fournir une première recommandation d'action d'une action à un jour sélectionné pouvant être déduit de données de recommandation d'analyste publiées, de graphes de corrélation croisée et de réseaux de corrélation partielle, une unité de modèle de série chronologique de comportement d'action pour déterminer un mouvement ou un comportement de prix d'action prédit de l'action pour le jour sélectionné à l'aide d'un modèle de série chronologique, une unité de prédiction de comportement multimédia pour déterminer un score d'humeur multimédia de l'action à l'aide d'un modèle de série chronologique d'humeurs multimédia, et une unité d'agrégation pour fournir une seconde recommandation d'action de l'action par agrégation de la première recommandation d'action, du mouvement ou du comportement de prix d'action prédit et du score d'humeur multimédia à l'aide d'un modèle statistique prédéfini.
PCT/MY2020/050059 2019-09-27 2020-07-27 Système et procédé permettant de fournir une recommandation de marché boursier WO2021060968A1 (fr)

Applications Claiming Priority (2)

Application Number Priority Date Filing Date Title
MYPI2019005734 2019-09-27
MYPI2019005734 2019-09-27

Publications (1)

Publication Number Publication Date
WO2021060968A1 true WO2021060968A1 (fr) 2021-04-01

Family

ID=75165961

Family Applications (1)

Application Number Title Priority Date Filing Date
PCT/MY2020/050059 WO2021060968A1 (fr) 2019-09-27 2020-07-27 Système et procédé permettant de fournir une recommandation de marché boursier

Country Status (1)

Country Link
WO (1) WO2021060968A1 (fr)

Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2012079227A (ja) * 2010-10-05 2012-04-19 Toshiba Corp 注目対象予測装置
KR20180117800A (ko) * 2017-04-20 2018-10-30 주식회사 비아이큐브 자산 포트폴리오 추천 방법
KR20190091743A (ko) * 2018-01-29 2019-08-07 주식회사 이상투자그룹 주식투자 정보제공 시스템 및 이를 이용한 추천종목 서비스 방법
WO2019171492A1 (fr) * 2018-03-07 2019-09-12 株式会社日立製作所 Dispositif et procédé d'assistance de tâche de prédiction

Patent Citations (4)

* Cited by examiner, † Cited by third party
Publication number Priority date Publication date Assignee Title
JP2012079227A (ja) * 2010-10-05 2012-04-19 Toshiba Corp 注目対象予測装置
KR20180117800A (ko) * 2017-04-20 2018-10-30 주식회사 비아이큐브 자산 포트폴리오 추천 방법
KR20190091743A (ko) * 2018-01-29 2019-08-07 주식회사 이상투자그룹 주식투자 정보제공 시스템 및 이를 이용한 추천종목 서비스 방법
WO2019171492A1 (fr) * 2018-03-07 2019-09-12 株式会社日立製作所 Dispositif et procédé d'assistance de tâche de prédiction

Non-Patent Citations (3)

* Cited by examiner, † Cited by third party
Title
SAKAMI KOUSUKE: "B1-4 News article analysis and its application to investment support", 17TH ANNUAL MEETING OF THE DATABASE SOCIETY OF JAPAN, DEIM FORUM 2019, 19 April 2019 (2019-04-19), pages 1 - 5, XP009527080 *
TAKAMINE, WATARU; IZUMI, KIYOSHI; SAKAHI, HIROKI; MATSUSHIMA, HIROYASHU; SHIMADA, TAKASHI; SHIMIZU, YASUHIRO: "Automatic Summarization of Analyst Reports Based on Causal Relationship Text-Mined from News Reports", SPECIAL INTEREST GROUP ON FINANCIAL INFORMATICS (SIG-FIN), no. 022-11, 5 March 2019 (2019-03-05), pages 1 - 5, XP009527057 *
TOSHIYA KATAGI, HIROKI SAKAJI, IZUMI, YASUSHI ISHIKAWA, KOHEI KASAOK: "Stock Price Analysis in Mid-to-long Term Using Analyst Reports", SPECIAL INTEREST GROUP ON FINANCIAL INFORMATICS(SIG-FIN), 5 March 2019 (2019-03-05), pages 1 - 6, XP009526953 *

Similar Documents

Publication Publication Date Title
US8069101B1 (en) Financial methodology to valuate and predict the news impact of major events on financial instruments
US20200192894A1 (en) System and method for using data incident based modeling and prediction
US11257161B2 (en) Methods and systems for predicting market behavior based on news and sentiment analysis
CN109859052B (zh) 一种投资策略的智能推荐方法、装置、存储介质和服务器
US20200151565A1 (en) Applied Artificial Intelligence Technology for Processing Trade Data to Detect Patterns Indicative of Potential Trade Spoofing
US8930254B2 (en) Financial methodology to valuate and predict the news impact of major events on financial instruments
Liu et al. A momentum threshold model of stock prices and country risk ratings: Evidence from BRICS countries
US20220343433A1 (en) System and method that rank businesses in environmental, social and governance (esg)
US8694413B1 (en) Computer-based systems and methods for determining interest levels of consumers in research work product produced by a research department
Sang et al. An in-depth analysis of the Altman’s failure Prediction model on corporate financial distress in Uchumi supermarket in Kenya
Krishnan et al. Are capitalized software development costs informative about audit risk?
Ganatr et al. Spiking back propagation multilayer neural network design for predicting unpredictable stock market prices with time series analysis
Wibisono et al. The use of big data analytics and artificial intelligence in central banking
Jammazi et al. Estimating and forecasting portfolio’s Value-at-Risk with wavelet-based extreme value theory: Evidence from crude oil prices and US exchange rates
Sharma et al. Are FinTech, Robotics, and Blockchain index funds providing diversification opportunities with emerging markets? Lessons from pre and postoutbreak of COVID-19
Chen et al. Understanding the impact of trade policy effect uncertainty on firm‐level innovation investment
Kirkos et al. Audit‐firm group appointment: an artificial intelligence approach
WO2021060968A1 (fr) Système et procédé permettant de fournir une recommandation de marché boursier
SirElkhatim et al. Prediction of banks financial distress
Bai et al. Offshore sales networks and stock return predictability
KR20230062256A (ko) 기업정보를 이용한 기업 투자정보 제공 시스템
CN113850609A (zh) 一种客户管理系统、方法、计算机设备和存储介质
Minnoor et al. Nifty price prediction from Nifty SGX using machine learning, neural networks and sentiment analysis
Jeon et al. Building industry network based on business text: corporate disclosures and news
Khandaker et al. Volatility and co-movement: an analysis of stock markets behavior of selected emerging and developed countries

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: 20869027

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: 20869027

Country of ref document: EP

Kind code of ref document: A1