KR100410714B1 - financial engineering based financial assets price forecasting system - Google Patents

Abstract

The present invention is a financial asset price prediction information providing system using wired and wireless Internet, and more specifically, the future price of financial assets such as comprehensive stock index, industry stock index, individual stock price, foreign stock index and stock price, interest rate, exchange rate, In particular, it predicts the price of 1st, 2nd, and 1st periods of the short term, and provides the information as a graph in 1 minute unit using wired / wireless internet, e-mail, client server, mobile phone, mobile stock terminal, and portable computer (PDA). It is a system. The present invention improves the accuracy by providing the price of the financial assets of the future (the short term) in a specific number, and provides accurate forecast information by providing the forecast information in detail in one minute, thereby significantly reducing the price of the financial assets that will occur in the future. By predicting and providing in advance with probability, it provides useful value information to investors of financial assets, and provides forecast information in real time, providing prompt information to investors and users of information trading in real time. In order to improve the usefulness of information by providing efficiency in judgment, and providing predictive data in specific numbers, it is possible to have a much faster understanding of information by providing predictive information in a graph.

Description

Financial engineering price forecasting information providing system using internet {financial engineering based financial assets price forecasting system}

The present invention is a system for providing financial asset price prediction information using the Internet, and an object of the present invention is to provide information to users by predicting the future price of an individual stock price, in particular, the first and second days, the first and second days of the short term. Promote and provide specific information by predicting price movement in one minute, and provide accurate information by providing price as a single numerical value instead of being provided in text. By presenting it as a plan, it is possible to provide easy information (promoting convenience), and by using the Internet and mobile communication means (wireless internet) as a means of providing information, anyone can use it quickly and easily.

The object of the present invention is not only for the financial assets, but also applied to the financial products of the global financial market, such as the United States, Japan, Europe, etc. can provide forecast information, and the information users are not only targeted to domestic people, global financial asset investors The same can be applied to.

The present invention relates to the prediction of the price to be formed in the future of a financial asset. More specifically, the price of a financial asset to be formed in the future to the user, that is, the stock price of an individual item, the domestic comprehensive stock index, the industry stock index, and the foreign composite In order to help users understand in a minute value by predicting the price to be formed in the future, especially in the short-term, in the short term in the future, especially in the short-term, such as stock index, industry stock index and stock index futures. As a graph, forecast information is provided in real time in order to reflect the ever-changing information for the purpose of promptness, and various information is provided through the Internet, e-mail, user server, mobile phone, mobile stock terminal, etc. to help users' convenience. It is a technology that provides an interface to easily acquire and access information.

The basic algorithm of the predictive model used in the present invention is based on the change rate of the minute stock price and the transaction volume that were formed recently (1-5 days) as the input variables. This is a method of finding a date with a high similarity to that of, and predicting the price to be formed after the current point in a minute based on the price movement formed thereafter. In the process of finding similarity, the neural network model learning theory is used, and the resultant formalized pattern is a method of finding basic patterns of stock price flow in chaos theory. There is no service to predict and provide graphs.

Currently, services performed by securities companies and stock information providers are mainly limited to providing current prices, moving average prices, and technical indicators of securities. Since all of this information uses the stock price or trading volume realized in the past, it does not provide information on the specific flow of the future stock price, but merely provides a buy and sell signal at the present time. In addition, the existing technical analysis only predicts the target price range prediction or the price rise or fall prospects, and does not visually predict the flow of the price, and thus there is a significant difference from the result of the prediction shown in the present invention. In addition, the current stock price analysis using technical indicators such as moving average line, separation, investment psychology, balance sheet, MACD, RSI, and CCI reflects much of the analyst's subjective thinking. Frequently, information users can be confused. Algorithms of technical analysis already developed cannot be treated as special techniques because they are mostly simple calculation processes as presented in textbooks. However, the present invention is based on the neural network theory and chaos theory, which are engineering theories. It is evaluated that it requires highly academic knowledge that cannot be compared with existing securities information service.

In addition, even though technical analysis using minute data is used in various methods of analyzing the current stock price, it is difficult to see it as accurate minute prediction because it does not accurately represent future prospects by minute. Stock prices are being forecasted by securities analysts, but mid- to long-term forecasts are mainstream, because the current and future time gaps (gaps) are so large that they do not quickly reflect rapidly changing information. In other words, there is a need to revise the forecast according to the reflection of information, and there may be ambiguity (no prediction) about the intermediate process (intermediate period) between the present and the forecast period, which may cause confusion for investors and related parties.

In view of the problems of the prior art, the present invention has developed a technology for predicting and providing a price of a financial asset to be formed in the near future and the most needed by investors and related parties, while reducing errors caused by changes in information.

In addition, forecast information to date presents wide boundaries and presents them in text, which causes more confusion for users. Therefore, the present invention has improved the accuracy by providing the price of the financial assets of the future (ultra-short term) in a specific number, and has developed a technology for providing accurate forecast information by providing forecast information in detail in one minute.

On the other hand, the forecast information of financial assets is presented at random times by experts to change trends and accidents at special times, thus limiting the timeliness of providing information to users (investors and related parties). For this reason, PACE provides users with timely information by quickly, continuously and regularly in five-minute increments (the fastest or most formalized time among 1 minute, 2 minutes, 5 minutes, and 10 minutes) to reflect the change in information. We have developed a technology that we can provide.

On the other hand, it has recently emerged as a new medium for information delivery, and is easily and automatically available anytime, anywhere through various interfaces such as the Internet, e-mail, client server, mobile phone, mobile stock terminal, and PDA. We have developed a technology that allows users to easily access information and access information.

1 is a view showing the overall service configuration of the prediction system and the flow of the prediction algorithm,

2 is a view showing in detail the configuration diagram of the present invention centered on a supplier server,

Fig. 3 is an interface of the present invention. Fig. 4 is a flowchart of the prediction algorithm of the present invention. Description of the main parts of the drawings 200: quote server 210: DB server 220: calculation server group 230: WEB server 240: wireless server 400: Price input unit 410: relevant asset price pattern learning unit 411: similar asset price pattern learning unit 420: learned pattern and corresponding asset price comparison / analysis unit 421: exogenous variable analysis unit 422: weight calculation 423: scenario adequacy review 430: expected Scenario calculation unit 440: Forecasting result DB450: Estimated yield calculation unit 460: Estimated yield DB470: Top estimated yield search unit 480: Recommended item selection unit

According to an aspect of the present invention for achieving the above object,

Predictive information such as composite stock index, stock price of individual stocks, stock index, stock index futures, etc. after 1 minute, 10 minutes, 1 hour, 1 day, ... In detail, the forecast price is suggested in 1 minute increments. In addition, it is not provided as a numerical value for the convenience of the user's understanding, but is provided as a graph so that it can be recognized at a glance, and the information is provided in real time to promote the speed of the information.

It provides the future prices of financial assets such as the composite stock index, stock price of individual stocks, stock index, and stock index futures, which are the targets of forecast information, in specific figures and prices.

The future prices of financial assets, such as the comprehensive stock index, stock price of individual stocks, stock index, and stock index futures, which are the targets of forecasting information, are provided as a graph for the convenience of users.

The future price of financial assets such as the composite stock index, stock price of individual stocks, stock index, and stock index futures, which are the targets of forecasting information, is provided in real time.

The present invention for achieving the above object, the step of connecting to the user's homepage through a web browser and a web server as an Internet connection means, and sequentially opening a plurality of input windows of the homepage to select the user's prediction information And opening the corresponding information file according to the user's selection and providing the user with forecast information on the relevant financial asset.This information is composed of only the user's server even though the user's server is not connected to the Internet through the user's server. When connected, the forecast information can be obtained in real time at any time. Even if the user does not directly access the homepage, if the user selects the information item of interest, it is provided to the user via e-mail every day. The forecast information is available through mobile phones and mobile stock terminals as well as the Internet. Consists of secondary steps to provide The.

Hereinafter, a preferred embodiment of the present invention will be described in detail with reference to the accompanying drawings.

-First, the present invention generates forecast data such as the composite stock index, industry stock index, individual stock price, foreign stock index and stock price, stock index futures, interest rate, and exchange rate (calculation algorithm).

-Build this in the DB of the supplier server.

-Upload the forecast data to the supplier's server's homepage web server.

-Prediction data stored in web server is provided to user through internet homepage access.

-Prediction data stored in the web server is automatically provided to the user's e-mail account, even if the user does not access the website of the Poce via e-mail.

-Prediction data stored in web server is provided to user through user's client server.

-Prediction data stored in the web server is provided by accessing the user's mobile phone.

-Prediction data stored on the web server is provided to the user's mobile stock terminal.

-Prediction data stored in web server is provided to user's PDA.

The operation of the present invention having the above function will be described briefly in the accompanying drawings.

1 is a flowchart illustrating a flow of acquiring information suitable for a user's request by selecting a user from a member authentication and menu selection by logging on in a flowchart of the present invention.

Fig. 2 is a block diagram of the present invention in which past prices of financial assets are graphed and converted into future forecast prices via the claimant's system. In the forecasting system of the supplier server, 1. Process the stock quote raw data. A calculation server for performing prediction algorithms using the raw data processing section, a calculation server for performing the prediction algorithm, and a DB server storing the prediction results calculated by the executed algorithm in a DB form, and performing WEB server that stores the prediction result calculated by the calculated algorithm in the form of a file, 5. A wireless server that transmits the prediction result calculated by the performed algorithm through the wireless network, and 6. Sends the prediction result to each user. Looking at the details of the provider server, first, the DB server 210 in the prediction system for the service is a raw data from the price DB (200) that provides a stock quote, which is the raw data (Raw Data) After receiving and storing a in a DB form, the raw data 251 transformed into a file form is transferred to the calculation server group 220, and the calculation server group uses the provided raw data file. Implement a forecasting algorithm that predicts the stock price of each individual item. The prediction result calculated by performing the prediction algorithm is generated as a DB type prediction result and a file type prediction result, and each type of prediction result is stored in the DB server through the DB server storage path 253 and the web server storage path. The stored prediction result is directly requested by the wired user and the WEB server 230 which is requested to transmit the prediction result transmits the prediction result to the wired user in real time. Accesses the wireless server 240 and requests the transmission of the prediction result, and the requested wireless server 240 requests the DB server 210 to transmit the prediction result of the corresponding item. The requested DB server 210 transmits the prediction result of the corresponding item to the wireless server 240, and the wireless server 240 transmits it to the wireless user again so that the wireless user can see the prediction information. Is shown in the form of pre-constructed by the user's prediction service interface.

3 is a diagram showing an interface through which a predicted price of a financial asset in the future is reached by a financial asset price prediction information providing system according to the present invention. FIG. 4 is a flowchart of a prediction algorithm. The main flows of the forecasting, index prediction, futures prediction, stock price prediction, interest rate prediction, and exchange rate prediction algorithm are as follows: First, the process of collecting raw data of the financial asset and converting it into raw data in the form of a file (400). After inputting into the price data variables in the program to execute the prediction algorithm, the neural network is applied to the price structure of the asset for the past 10-15 years using the price information formed on the last 1-5 days received as input variables. Analyze the correlation through the similar pattern learning process (410) applying the theory, and sort the results by the highest correlation (based on R-square value) (the date, time, Catastrophic) is stored in a temporary variable. In addition, to compare prices with similar assets, the price structure of similar assets in the past 10-15 years (for the Composite Stock Index, similar assets are kospi200 and kospi200 futures futures, and for individual stocks, similar assets are industry representative stocks). Analyze the correlation through the similar pattern learning process (411) applying neural network theory, and store the result (the date, time, price change) as the temporary variable with the highest correlation (based on R-square value). Based on the result of the price analysis of the asset and the similar asset, in the process of comparing / analyzing the learned pattern with the price of the asset (420), the top three results of the price analysis of the learned asset and the top three results of the similar asset price analysis Find the corresponding date in the past where the result was selected, calculate the rate of return formed thereafter, and then apply the rate of return after that, based on the present price, to the expected price of the asset to be formed after the current point in time. After that, it is stored as a temporary variable (expected price time series 1,2,3,4,5,6) .In the exogenous variable analysis step, the comprehensive stock index, the industry index, and the trading volume of the financial asset are used as exogenous variables. Analyze correlations through exogenous variable analysis process (421) applying neural network theory to data structures of exogenous variables in the past 10-15 years as input variables using information of exogenous variables formed in the last 1-5 days received. By sorting the order of exogenous variables by the highest correlation (based on R-square value), find the relevant date in the past when the result is selected, calculate the rate of return formed thereafter, and apply the rate of return after that based on the current price. Estimate the estimated price of the asset to be formed later and store it as a temporary variable (expected price time series 7,8,9); similar to time series data (1,2,3) predicted from historical pattern analysis of the asset; Past loss of assets Based on the rate of change of nine time series data (4,5,6) predicted from the analysis and time series data (7,8,9) predicted from the past pattern analysis of exogenous variables, the chaos theory is applied. In order to minimize the variance of the weighted average returns, the constraint is that the optimal weights of each time series are obtained by solving a nonlinear equation in which the weighted average return of the predicted time series is equal to the simple mean return and the sum of the weights is 1. The weight selection 422 of the analysis results of the asset and the exogenous variable includes the programming of the above optimization model. The review of the appropriateness of the temporary scenario 423 includes all of the following processes. The correlation coefficient between the weighted average predicted rate of return and the time series of the individual predicted rate of return, which apply the nine optimal weights found in the six-step weighting analysis of the asset and exogenous variable analysis (422), is estimated. If there are three or more time series, the prediction scenario is determined to be appropriate, and the price data and the charts of the corresponding time series are implemented in the prediction graphs 1,2,3 in order of high correlation coefficient. If there are less than three time series with a correlation coefficient of 0.5 or more, the forecast scenario is judged to be inadequate, and the process returns to step 4 of comparing and analyzing the learned pattern and the corresponding asset price (420). Based on the results, repeat steps 4-7 until it passes the adequacy criteria. Once the appropriate scenario is calculated, it is stored in the DB of the prediction result of the DB server. Based on this, the process of selecting the recommended items in the calculation server is as follows. The recommended items are selected based on the prediction results stored in the DB. The process of calculating the expected return of the predicted items in the stored DB (450)-Prediction The forecasted market price and forecast closing price are extracted for each scenario, and the expected return for each scenario is calculated. By calculating the average expected return by combining this rate of return-the return of the forecast scenario for each item is calculated and stored (460), and the result of the average expected return for each item through the upper expected return search process (470) has a high expected return The ranking is determined in order, and is expressed in the form of item code, item name, forecast time point and expected average return. The result is provided in real time through each path at the request of wired and wireless users.

As described above, by providing a predictable value of the future financial asset price with a considerable probability, it provides useful value information to the financial asset investor and provides forecast information in real time to deal with the sale in real time. Providing prompt information to investors and users of information to improve efficiency in investor's investment judgment, and provide useful information by providing forecast data in specific figures. It aims at accuracy, and the prediction information is provided as a graph, which makes the information much faster.

In addition, while described in detail with respect to the preferred embodiment of the present invention, those of ordinary skill in the art, the present invention without departing from the spirit and scope of the invention defined in the appended claims It will be appreciated that the various modifications or changes may be made. Therefore, changes in the future embodiments of the present invention will not be able to escape the technology of the present invention.

Claims (8)

Stock quote processing unit for processing the basic data of the stock quote, algorithm performing section for performing the prediction algorithm using the basic data, database server for storing the results calculated by the performed algorithm, Web storing the prediction results in the form of a file Using the financial asset price information providing system including a server, a supplier server for processing and transmitting the prediction result through a wired or wireless network, Collecting basic data of the financial asset and converting the basic data into a file format and inputting the data as a quote data variable of the prediction program (first step); Asset price pattern learning step of analyzing the regular price information price structure through neural network theory and chaos theory (second step); Learning the similar asset price pattern by comparing / analyzing the price data variables and the price structure of similar items through neural network theory and chaos theory (3rd step); Generating an expected scenario using the corresponding asset price pattern analysis result and the similar asset price pattern analysis result (step 4); Analyzing exogenous variables present in the market (step 5); Generating a weight using an exogenous variable (sixth step); Examining the appropriateness of the expected scenario by applying the expected scenario at a time point after the actual learning (step 7); If the result of the suitability review (step 7) is inappropriate, performing steps 4 to 7 again (step 8); Calculating the expected rate of return for a plurality of items (step 9), if the result of the suitability review (step 7) is appropriate; Selecting items with high expected returns as recommended items and generating a list of recommended items (step 10); Providing minute-by-minute prediction graph information for each item at the request of a wired or wireless user (step 11); Financial asset price prediction information providing system using the Internet, comprising a delete delete delete delete delete delete delete