TW201222455A - Method, computer program product and computer-readable recordable medium for selecting investment target - Google Patents

Abstract

The invention relates to a method, a computer program product and a computer-readable recordable medium for selecting an investment target. The method includes grouping all investment targets within a first time slot based on whether the performance of the industry of the investment targets beats the performance of index, whether the performance of any individual investment target beats the performance of the industry belonged thereto and whether the performance of any individual investment target beats the performance of the index within the first time slot. The method further includes estimating a pattern within a historical time slot and calculating the odd of going up and the average gain, the odd of going down and the average loss and the odd of being flat based on the specific grouping pattern of some specific investment targets within a second time slot.

Description

201222455 VI. Description of the Invention: [Technical Field of the Invention] The present invention relates to an investment 潠Ηπ_ 士 only negative selection method, a computer program product, and a computer readable recording medium, in particular, such as ^^J疋扎Standardized investment stock selection methods, computer program products and computer readable recording media. [Prior Art] The scope of the securities or financial investment industry is + _, and the proportion of people engaged in securities investment is also quite high. However, there are not many investors trained in professional investment knowledge. And those who have been trained by the # knowledge, or the so-called information intermediary (including investment, investment or stock market name) are based on professional conscience and stand in the corner of the investor ^ 'to seek the maximum profit or reward of investors as the starting point And to convey pragmatic business operations or market information to the investing public, it is not known and 'it is not timely to verify the correctness of the information provided by MXC, or to use scientific methods to (4) its recommended stocks_ Accuracy|usually makes u many investors become the losers of this market, although market investors or professional investors such as analysts are used to: often do the stocks relative to the industry or the market information In comparison, however, there is no related technology to incorporate this three kinds of information for stock selection. In addition, when evaluating the stock stock selection benefits, even if investors choose to sell better than the industry, they may not be better than the broader market, or the stocks that are better than the broader market may not be better than the industry. Even the squatting is based on the best of the three market considerations, and there is still no guarantee that it will not be guaranteed. Therefore, it is necessary to seek a solution 201222455 [Summary of the Invention] Therefore, 'the purpose of the present invention' is to provide a standard investment stock selection method, so that the invention standard stock investment stock selection method for an investor to select at least one suitable investment Subject. The standard investment stock selection method comprises the following steps: (A) according to a stock market price database, calculating historical reward data and historical historical reward data of all stocks in all first time intervals including a historical time interval of a recent time, And the historical remuneration data of each industry classification; (B) According to the historical time interval, whether the historical remuneration of the stocks of the stocks in all the first time intervals is greater than the historical remuneration of the large-capital stocks, and whether the historical remuneration of the individual stocks is greater than the historical remuneration of the industries to which the individual stocks belong, And whether the historical remuneration of individual stocks is greater than the historical remuneration of the market, all the stocks are grouped in all the first time intervals to generate the group of all the stocks in all the first time intervals ^ and (6) according to the - special stocks in - including the In the second time interval of the most recent time interval, the material grouping mode belongs to (4), and the same specific grouping mode appears in the historical time interval, and then the first one time is the rising probability and the rising average salary of the domain, the history The same specific clustering mode occurs within time (4) and the subsequent - first-time interval is falling The probability of falling and the average return of falling, the specific sub-chain in the historical time interval is the same as 4·' α # νζ» ® I. The first time interval is the flat disk probability of the flat eight. Probability and flat-panel -, rising probability and falling probability for the investor to use for the stock. It is a further object of the present invention to provide a computer program product. 201222455 The program is loaded via a computer. Thus, the computer program product of the present invention performs the above-mentioned standard investment stock selection method. A further object of the invention. That is, a computer readable recording medium is provided. Thus, the computer readable recording medium of the present invention records a code which can be controlled by the brain, and the code includes the execution steps of the above-mentioned standard investment stock selection. & method

The effect of the invention lies in that the primary information published by the stock exchange is deeply explored, and the results of the comparison of the two and two rewards are used to gradually classify, and the market performance is similarly classified, and then the prior information is used. The historical performance of the same stocks is further subdivided, gradually narrowing the range of stock selection, and then finding the probability of future stocks rising and falling and the expected repayment value' thus providing investors with valuable and valuable information. [Embodiment] The above and other technical features of the present invention will be apparent from the following detailed description of the preferred embodiments. Before the present invention is described in detail, it is to be noted that in the following description, similar elements are denoted by the same reference numerals. Referring to Fig. 1, a preferred embodiment of the standard stock investment method of the present invention is implemented by a standard stock investment stock selection system 10 placed in the form of a software of a system side server 1. The standard investment stock selection system 1 is for an investor to select at least one suitable investment target (such as a listed stock or a portfolio), and includes a stock market compensation data computing module 11, a grouping module 12, The grouping module J 3 201222455 is a probability and average reward computing module 14, a risk expectation reward computing module 15, a decision and screening module 16 and a regression computing module 17. The stock market compensation data calculation module is used to calculate the stock price database 8 at the other end of the Internet 9 (such as the stock price data provided by the Taiwan Stock Exchange), including all the historical time intervals of a recent time interval. Historical remuneration data for all stocks in the first time interval, historical remuneration data for large-cap stocks, and historical remuneration materials for each industry classification. In a preferred embodiment of the present invention, the historical time interval is from January 1, 1971 to March 1, 2010, the first time interval is one, and the most recent time interval is March 2010!曰 (or the same day, in the following preferred embodiment, the daily data of Taiwan listed counter stocks from January 1, 1971 to March 1, 2010 is taken as an example. However, in this In other embodiments of the invention, the first time interval is not limited to one, but may be a larger or smaller time unit, such as 1 hour or 5 minutes, etc. The grouping module 12 is configured to be within the historical time interval. Whether the historical remuneration of the industry in which all stocks in the first time interval belong is greater than the historical remuneration of the market, whether the historical remuneration of individual stocks is greater than the historical remuneration of the stocks in which the stocks belong, and whether the historical remuneration of individual stocks is greater than the historical remuneration of the stocks, and all of them will be The first time interval is grouped to generate the belonging groups of all the stocks in all the first time intervals. That is, the group module 12 is based on the industry and the market, and all the stocks are all first. The time interval is grouped to generate 8 groups. As shown in Table 1 below, the performance of any listed stock at a certain point can be classified as mutually exclusive eight groups. 55 'Group 1 can be called a strong stock, group 8 can be called a weak stock table 1 industry> Large stocks> Industry----- stocks> Large group 1 --- Η Yes Yes Group 2 is whether group 3 is group 4 or not group 5 no yes -------^ is group 6 no group 7 no no---___, yes group 8 no No---J η τ clarifies the industry classification of Shangyu Cabinet Company, including the broadly divided 'middle and downstream vertical or horizontal sub-industries, related companies, concept stocks, group stocks (such as ETFs), depositary receipts (such as ADR, R * TDR), high revenues, or any type of stock that can be classified as common, to form an industry. In the preferred embodiment, the Bay Stock Exchange The stocks provided by the counter buying center (ie the stock exchange classification) are classified as industrial classifications. The industry model: according to the grouping mode, the grouping mode or the row. The so-called grouping time interval is the historical reward group. In the two time interval, the two shares belong to the group within the most recent time interval including the most recent time interval. Second: the number of specific groups and the historical reward data are divided. Hite 疋 individual stocks in the historical time interval set the knife group 'the ten second time interval can be, for example, 5 201222455 days. The so-called array grouping means The historical reward data of the specific individual stocks in the historical time interval is grouped according to the specific arrangement of the specific individual stocks in the consecutive time-time intervals in the second time interval including the most recent time interval. In the description of the method embodiment, the combination and arrangement of the two grouping modes in the present invention will be described in more detail. The probability and the average reward operation touch 14 cores (4) the group of the individual shares in the second time interval In a specific clustering mode, the same specific clustering mode occurs in the historical time interval, and the subsequent - first time interval is the rising probability of rising and the average rising salary, the falling probability of falling and the average falling return, and the flattening The probability of the disk, in which the sum of the probability of rise, the probability of falling, and the probability of flatness is 1. The risk expectation reward calculation module 15 is configured to calculate the position risk expectation reward and the unit expected reward risk of all stocks to form a risk expectation reward list 140 or 140a. The determining and screening module 16 is configured to filter out stocks of all the stocks that meet a predetermined screening criterion, and sort the selected stocks by the unit risk expectation rewards from large to small, or expect the unit to be The remuneration risk field is sorted from small to large to obtain a portfolio list of 16 or 16〇a for the investor to decide whether to buy or sell the stocks in the portfolio list 16〇 or 160a. The regression module 17 uses a complex regression model to calculate individual stock returns. Referring to Figures 1, 2, 6, and 7, the initial step 30 of the preferred embodiment of the standard stock picking method of the present invention in the combined grouping mode is automatically based on the stock market compensation module 201222455. The stock market remuneration database 8 calculates the historical remuneration data of all stocks, the market and the classification of each industry from January 1, 1971 to March 1st. Then, as shown in step 31, the grouping module 12 according to the historical time interval of the period from January 1, 1971 to March 2010, whether the historical reward of the industry in which each stock market is operated is greater than the historical reward of the market. Whether the historical remuneration of individual stocks is greater than the historical remuneration of the individual stocks, and whether the historical remuneration of individual stocks is greater than the historical remuneration of the broader market, all stocks are grouped to generate all the stocks in each group. As shown in step 32, the grouping module 13 is a specific group of eight groups according to a specific stock in a second time interval including the most recent time interval (for example, the number of groups and the latest time) The group within the interval, the historical reward data of the specific individual stocks in the historical time interval is combined and grouped. It is particularly pointed out that steps 32, 33, and 36 in Figure 2 are sequentially performed for a specific stock. For example, in the preferred embodiment, since the steps 32, 33, and 36 are performed for all the stocks in the order of the company code, the group mode 13 is first directed to the stock of the company (company code no). The traps 1411, 1412, and 144 in the platform company risk expectation remuneration list 140 shown in FIG. 6 are formed in the combined grouping mode for the counter mud stocks generated in step 31 and the other: Beinjin group. Figure 6 shows the risk expectation remuneration list (10), which lists the relevant information of the historical business day from February 6 to February 2 of the 2nd year of the use of the present invention. The block 1412 is the historical business day. Group of the day 'for example, March 1 The group to which it belongs is group 4. The block 1411 201222455 is the number of days belonging to the group in the first 4 days (that is, 5 days in total) for each historical business day, for example, in the case of March i, due to 3 On the 1st of February, February, February, February 25th, February 24th, and February 23rd, the business days of Group 1 for the five consecutive historical business days are February 24 and February 23 for 2 days. Therefore, the number of 1411 for the 1st day of the group on March 1 is 2. The field 144 is the number of days in the historical time interval that belong to the same grouping mode, for example, from January 1, 971 to 2010. In the historical time interval of the 1st of the month, it has the same combined grouping mode as the March 1st of 2_ years (that is, the day plus the first 4 days) and the other J 1 days is 2 days and the day group is the group 4) The number of days is 214 曰 ° Next, as shown in step 33, the probability and average reward calculation module 14 operates the same combination grouping of the stocks of the company in the historical time interval from January 1, 1971 to March 2010. The pattern and each day after the day is the upper (4) rate (block 145) and the loading _ (touch 142), the falling probability (field H6) and the falling average reward (field 143), In order to flatten the probability of flatness (not shown in Figure 6), the sum of the probability of rise, the probability of falling, and the probability of flatness are 丄, and the probability of rise and the probability of falling are for the voter. In order to determine whether to buy or sell the particular stock. As shown in Figure i, the investor can operate the memory device 22 of the client device 2 (such as a personal computer including the processing benefit 21, 5 memory 22 and display 23). The stock picking system operation "face program 22G is connected to the system side server〗 to display the risk expected reward list 14〇 generated by the standard stock selection system H) on the display system 23 of the display system 23 of the display 23 On the operation screen. In the preferred embodiment of the risk expectation remuneration list 14 of Figure 6, the system 10 201222455 uses discontinuous probability densities to calculate the relative number of shares on the second share of the same combined grouping mode. That is, in the present preferred = upper...

2 Remuneration Group 14 is calculating the probability of the rise of the platform mud stock in the next camp on the second day of March i. (4), the first calculation of 214 history: the day ten mother - after the history of the day - the day is rising The number of days of the rise, and then the full rate on the 2nd of March is equal to the number of days of the increase divided by the total number of days (that is, the number of the same group). In the same way, the probability and average reward calculation module is in the process of calculating the falling probability of the platform mud stock on the next business day of March 2, the second business day, the first time in the history of 214 historical days. The day after the fall is the number of days of decline, and the probability of a fall of March 2 is equal to the number of days of the fall divided by the total number of days (ie 214 days in the same group). Further, in another embodiment of the present invention, the probability and average reward operation module 14 may also use a continuous Probability Density Function (PDF) belonging to historical reward data to calculate the probability of rise and fall. That is to say, for the probability distribution of historical stocks of individual stocks in the same grouping mode, the hypothesis is normal allocation, and the stocks in the same grouping mode are in the recent historical time interval (such as March 1 in Figure 6)— The probability of stock return JC for a business day is Prob (x>=c), and the PDF of jc is as shown in the following formula (1): /0) = ^ϊπσ

(1) where μ is the average of jc and σ is the standard deviation of JC. In addition, if X is a common lognormal assignment in practice, its PDF is as shown in the following formula (2): 11 201222455

Jc> 〇 Tailu BH Μ ^(6) The average number of ports is the standard deviation of llU. Furthermore, in the embodiment, the probability and average reward calculation module 14 can also calculate the upper _ rate and the falling probability. The so-called conditional probability 疋 曰 事件 under the condition that the known event E occurs, the event. The probability of occurrence is p(g\e)=^J2^1 P(E) (3) P(E)> 〇G and E are independent of each other' corpse illusion: (10)*(10) corpse (brown) = (10). In addition, in still another embodiment of the present invention, the probability and average reward operation module 14 can also use the Bayers' Decision Rule to account for the probability of falling and the probability of falling. The so-called shell-like decision rule assumes that the sample two can be divided into r mutually exclusive events such as Gi to Gr, and the pre-existing probability p(Gj), j=l...n is known. Under the consideration of event ε, Find the probability ρ(Ε), such as the formula, and then according to Bayes' theorem, find the probability P(Gj|E) afterwards, as in formula (5). P(E)=P(E n G,)+ P(E n G2)+.....+P(EnGn) =PiG^PiElG^ P(G2)*P(E\G2)+.. = Tp(G^*p(E\Gk) k=! P(n\F} = ^.JnE^ Ρ^ΤΡ{Ε\0}) J 1 J pi ρ\ n yP(GkrP(E\Gk) + P(Gn)*/> (£ IGJ (4) (5) Next, as shown in step 36, the risk expectation reward calculation module 15 calculates the unit risk expected reward for a particular stock. In the preferred embodiment, The detailed process of this step 36 is to calculate the expected remuneration of a particular stock first (list 12 in 201222455). = rising probability, parent rising average return + falling probability, falling average remuneration, and then calculating the unit risk expectation of a particular stock. (Listing 149 of Listing 14) - Expected compensation divided by risk value, where the risk value can be calculated as the total risk, system risk, or standard risk, as shown by the standard deviation (Figure 147, stop 147) Risk value (Value Ai (4), var). In addition, the risk expectation compensation operation module 15 can also step into a specific stock list.

The expected salary is assumed to bear the risk = 1 / unit risk expected salary, and the unit expects the compensation risk field 150. Then, as shown in step 37, the decision and screening module 16 determines whether the rising probability, the unit risk expected reward, and the unit expected remuneration risk of all stocks have been obtained. If the determination result is no, as shown in step %, the standard investment stock selection system 1G transfers the operation object to the next specific stock, and then proceeds to steps 32, 33, 36 again, until all the stocks are obtained. The probability of rising, the expected return of unit risk, and the risk of the expected return of the unit. Therefore, when the result of the determination in step 37 is YES, then as shown in step 39, the determination and screening module 16 selects the screening criteria according to the group including the plurality of _reference data, and selects all the stocks that meet the predetermined screening of the group. Individual stocks of the Code. The screening criteria of the group related to the screening reference series can be set as the expected remuneration is positive (or the expected remuneration is negative short), the rising probability is greater than - the predetermined probability threshold 〇.5 long (or the rising probability is less than a predetermined probability threshold) The value is 0·", so the standard deviation is less than a predetermined probability gate value, and the base 2 is as shown in Table 2) 'The price-based technical indicators or the internal indicators (as shown in Table 3), each The pen averages, financing bills, stocks of the three major corporate entities, and the status of the chips, etc. As shown in Table 2 below and Table 3 of 201222455, they respectively show the complete fundamental financial ratio indicators and various technical indicators. It is used in the embodiment of the present invention. Table 2 Financial and financial structure ratio of credit (Note 1) Fixed assets to total assets ratio (fixed ratio, bank loan to net ratio, long-term debt to net ratio, long-term bank loans) Net worth ratio, fixed asset to net worth ratio (fixed ratio), fixed asset to long-term capital ratio (fixed long-term suitability ratio), debt-to-net ratio (leverage ratio), solvency (flow) Ratio, quick ratio (acid rate note 1) Bank loan analysis of current assets multiples (interest guarantee multiple, interest guarantee multiple (plus refill 1) depletion + amortization), net cash flow to interest expense ratio of business activities, Net cash flow to total debt ratio of operating activities, net cash flow to total debt ratio of discretion, net cash flow from operating activities to short-term bank borrowing ratio, net cash flow to capital expenditure ratio of operating activities, depreciation of capital expenditure + Debt + Amortization Ratio Cash Flow _________ ______ Cash Flow Ratio, Cash Reinvestment Ratio --------- 14 201222455 Analysis (Note l) Other cash flows see the gold machine summer allowance ratio, net cash flow from operating activities Ratio of interest expense, net cash flow of operating activities*, rate of bank borrowings in net cash flow from operating activities, net cash flow to capital expenditure ratio of operating activities, net cash flow from discretion, other long-term and short-term debt Capability (Note 2) Gold Coverage Ratio, Long-Term Liabilities to Shareholders' Equity The interest multiplier, the surplus is the expenditure multiple (fixed expenditure guarantee multiple), the cash interest coverage or cash coverage ratio, the fixed expenditure guarantee multiple, the debt principal and interest compensation ratio, the long-term debt residual ratio, the shareholders' equity to the total assets Ratio (equity ratio), long-term debt to total assets ratio, long-term capital to fixed asset ratio, debt-to-equity ratio, debt-to-capital ratio, long-term capital investment return rate, short-term coverage ratio or short-term protection ratio, net non-funded project ratio, Current assets to total assets ratio, current assets turnover rate, flow 15 201222455 What is /*···» · " I--------, debt-to-inventory ratio, current liabilities-to-net ratio, liquidity Index, allowance for bad debt rate, settlement ratio, current liabilities to total debt ratio, fund and long-term investment to total assets ratio, fixed cost guarantee multiple, capital turnover rate, net return rate, long-term bank loan to net worth ratio, double leverage ratio Contingent-to-net ratio, long-term capital suitability rate, borrowing dependency, allowance for bad debts, lending assets Debt discount + depletion ratio on gross depreciation ratio, cumulative industry analysis (Note 1) Depreciation on gross assets ratio, capital expenditure on fixed assets, gross ratio, capital expenditure on fixed assets net ratio Performance (Accounts Payable Turnover, Accounts Receivable Turnover, Inventory Note 1) Turnover Rate, Fixed Asset Turnover, Total Asset Turnover, Net Value Turnover, Net Working Capital Turnover Operating Efficiency (Stock Turnover Days) , average cash withdrawal period, accounts payable than asset utilization / payment days, deposit and conversion period, cash cycle turnover rate turnover rate, average cash cycle days, business cycle, net camp (Note 2) industry cycle, allowance Account rate, total asset growth rate, solid 16 201222455 疋 greedy growth rate, net value to assets, land to net value, fixed assets to net profitability (gross profit margin, operating profit margin, operating profit margin (deduction of interest note 1) fee ), net profit rate (before tax), net profit rate (after tax), net rate of return (tax), net rate of return (after tax), total Return rate of shell products (pre-tax, unpaid interest expense), total • rate of return on tributary (after tax, unpaid interest expense), total - return on assets (before tax, plus interest expense), total capital - Return on production (after tax, plus interest expense), depreciation + depletion + amortization to the people (four), interest expense to operating income ratio, other profitable inventory pairs • ability to run margin, net interest rate, sales return and discount Ratio, aspect financial sales discount ratio, lock cost ratio, operating ratio I ratio, operating expenses to net sales ratio, depreciation to net sales ratio, bad debt to credit sales ratio, sales (business) income pair Cash ratio, sales (business) income corresponding - collection ratio, sales (business) income to inventory ratio, sales 17 201222455 goods (business) income to current assets ratio) income to total assets ratio, sales t ( Business) income to working capital ratio, sales revenue to inventory ratio, lock-in revenue to working capital ratio, sales to net worth, sales revenue to net profit, net Proportion of sales revenue, net profit to working capital ratio, net profit to net worth ratio, pre-tax return on assets, after-tax return on assets, return on total assets, return on invested capital, income growth rate, net profit margin, net profit growth rate Net profit growth rate, pre-tax net (four), post-tax net (four), non-operating income and expenditure ratio, realized sales growth rate, operating margin growth rate, continuous net profit growth rate, operating profit to paid-in capital ratio, pre-tax net profit Earnings per share compared to investment-related ratios earnings per share, earnings per share 'per-tax net profit per share, cost-benefit ratio or equity ratio, dividend yield, market-to-book value ratio, earnings yield, surplus value index, principal Libby, dividend yield, dividend value index, net share value of parent stock, share price to net worth ratio, dividend payment or release rate, earnings retention rate, dividend growth rate, market price 18 201222455

Ί------ ... The ratio of cash flow, share price to cash flow per share, cash dividend protection multiple, cash flow per share, net worth growth rate, common stock return on equity, sustained growth rate Growth rate, re-return rate of earnings of common stockholders, return on pre-tax shareholders' equity, return on equity of pre-tax ordinary shareholders, return on equity after-tax, return on equity after-tax ordinary shareholders, return on long-term capital investment, Equity, Wuyi<τ ^ -h team knives total beben after-tax net profit rate, total capital pre-tax net profit rate, force shareholders' equity pre-tax net profit rate, after-tax net profit ratio, pre-tax net profit margin, gross profit margin' Marginal contribution rate analysis activity force Beben turnover rate, shareholders (Note 3), fixed assets turnover rate, operating equipment turnover rate, inventory turnover rate, accounts receivable turnover rate, net worth of fixed assets to fixed assets ratio), internal retention rate , solvency guarantee multiples, ability to repay long-term borrowing interest 'self-funding to capital expenditure ratio, thousand debt ratio, net interest burden rate, corporate blood pressure _ 19 20122245 5 Productivity added value, per capita added value, capital allocation rate, fixed asset productivity, equipment investment efficiency, per capita turnover, marginal contribution per person, human productivity, capital productivity, salary marginal revenue productivity growth financial institution special concept and Ratio Other financial concepts and ratios Note revenue growth rate, value-added growth rate, shareholder equity growth rate (net growth rate), pre-tax net profit growth rate, post-tax net profit growth rate, fixed asset growth rate spread, average lending rate , average deposit interest rate, problem assets, capital adequacy ratio, liquidity readiness ratio, overdue ratio, collection ratio, coverage ratio, bad debt ratio, storage ratio, non-interest income, operating income ratio, economic added value, free cash flow, difference Value (Z-Score), weighted capital cost, operating leverage, financial leverage, comprehensive leverage, financial leverage multiplier, a total of 45 financial ratios from the Financial Joint Credit Information Center of the Financial Corporation from other sources Note 3: Fully presenting a five-force analysis while retaining the same ratio as other sources f. 3____ Technical indicators species 20201222455

Invaluable technical indicators: Moving Average Convergence Divergence Indicators or Moving Average Convergence-Divergence (MACD), Directional Movement Index (MI), Stop Loss Point Steering Indicators. (Stop And Reverse, SAR), relative strength indicator. (Relative Strength Indicator ' RSI ), Deviation Rate (BIAS), William Index (Williams' %R, WMS%R), Momentum (MTM), Dynamic Momentum Index, Oscillator (Osi 1 lator) , OSC), Swing Index, Quantative Candle Stick (Qstick) 'Commodity Channel Index (CCI), trading momentum indicator or opening price ratio (AR), buying and selling willingness indicator or closing price Ratio (BR), Triple Exponentially Smoothed Average (TRIX), Contrarian Operating System (CDP), Stochastic Oscillator (KD), Stochastic Momentum Index, Random Relative strength indicator. (Stoch Relative Strength Index, Stoch RSI)' Relative Fluctuation 21 201222455 Index (Relative Volatility Index), Rate of Change (ROC), Psychological Line (PSY), Bollinger Band, Kent Keltner Channel, Absolute Breadth Index, Accumulation / Accumulation

Distribution line), Accumulation Swing Index, Williams' s Accumulation/Distribution, Advance/Decline Line, Advance/Decline Ratio , Median Price, Price Channel, Parabolic Stop and Reversal or Parabolic SAR, On Balance Volume (OBV), Cumulative Diffusion (Accumulation/ Distribution Line, A/DL), Money Flow Index (MFI), Volume Ratio or Volume Ratio (VR), Up: Falling Line (Advance Decl ine Line) (ADL) ), Average Volume (MQ), Volume Adjusted Moving Average (VAMA), Easiness of Movement (Ease of Movement, 22 201222455 EQM), Force Index (FI), each weighted share price Total Amount Per Weighted Stock Price (TAPI), Cai Jin Swing Indicator or Chaikin Osci 1 lator (CHKO) Chand Momentum Oscillator, Money Flow Relative Strength Index (MFRSI), Arms Index (AI), TRIN Short Term Trading Index (TRIN), Or Trader's Index, Stock Market Thrust (MT), Thrust Oscillator (TO), McClellan Oscillator (McClellan), McClellan, and Indicators (McCLELLAN Summation Index) ), Turnover ratio, Focus amount (MC), Volume indicator (Volume), Volume Osci 1 lator, Volume Rate-Of-Change, Positive indicator ( Positive Volume Index, Negative Volume Index, Equivolume, Price Increase, Price Fall and Price Constant Securities Volume (Advancing, 23 201222455

Dec 1 ining, Unchanged Volume), Upside/Downside Ratio, Upside/Downside Volume, Price and Volume Trend, Volume Moving Average (Advance Decline Ratio, ADR), Oversold Oversold (Bever/Over Sold, 0B0S), Counter-Clock (Counter-clockerwise), Eminem Stocks Index (ARMS Index), Stock Market Thrust (MT), financing balance, balance of securities and securities, balance of stock market cycle, computer-assisted transactions, number of transactions per five minutes, number of sheets and transaction value, the current day Reversal ratio. As shown in FIG. 7, 'the maximum reward for each unit of risk is the objective function'. In the preferred embodiment, when the predetermined screening criterion is the expected reward (field 166) is positive, the probability of rising (field 165) The value greater than a predetermined probability threshold of 0.5, the standard deviation (field 168) is less than 4, the revenue of the recent month (field 162) is positive growth, and the value of the 8th RSI (field 163) is between 15 and 60. The decision and screening module 16 will generate a portfolio list 160 as shown in Figure 7 wherein the definition of most of the fields in the list 160 is the same as that defined in Figure 6, for example, the block 164 refers to the nearest The historical business day (March 1 of March 1st) plus the number of days belonging to Group 1 in the previous 4 days (ie 5 days in total). The portfolio list 16 in this Figure 7 is sorted by the unit risk expectation reward field 167 to 24 201222455 and displayed on the stock picking (4) operation interface 230 of the display 23 of the investor device 2, so that the investor can be based on the shelf 167. (Or sorting the seats to select stocks that you want to buy and sell on March 2, 2010, such as buying the Berteng stock with the highest unit risk expectation 〇 628. See Figure 1 3 8 , 9 ' in the method of the invention In other embodiments, another (4) different grouping mode arrangement may be used to group the historical reward data of a particular stock in the historical time interval. As shown in FIG. 3, f is shown in other method embodiments of the present invention. Using the permutation grouping mode to implement the grouping of the historical remuneration data of the special stocks in the historical time interval. Since the steps in Fig. 3 are only steps 34 and 35 are different from the steps 2 and 33 in Fig. 2, the remaining steps are completely The same, so only steps %, 35 will be described in detail here. For steps 30, 3, and 36_39 in Fig. 3, reference may be made to the above description of the same steps 3, 31, and 36 39 in Fig. 2. Step 34 As shown, the grouping module 13 aligns the particular individual stock or all companies in the historical time zone according to the consecutive first time interval=group arrangement of the specific stocks in the second time interval including the = near time interval. The historical reward data is arranged and grouped. It is particularly mentioned that the steps 34, 35, and 36 of FIG. 3 are sequentially performed for a certain specific stock = for example, in the preferred embodiment, The code is for all stocks - step 34, 35, %, so the cluster module η first for the stock of Taimu (company code is 11〇1), in order to group all the company stocks generated in step 31 The group information is grouped to form fields 141 and 144a of the platform company risk expectation remuneration list 140a order shown in Fig. 8. 25 201222455 As shown in Fig. 8, the risk expectation reward list! The relevant information of the historical business day is from February 6th to March 1st of the Lion Year. The block 1412 is the group of the day on which the business day is, for example, the group belonging to U i曰 is a group. Don't mind blocking (4) for each calendar On the day of the business day plus each of the first 4 days (that is, a total of 5 days), for example, due to ... 曰, February 24, February 25, February 26, March 1, a total of 5 consecutive historical business days The group to which the day belongs is group 1 , group ^ group 8, group 5, group 4, so the value of the corresponding block 141 is the arrangement of 11854. The position correction is the historical time interval ten. The number of times the company arranges the grouping mode, for example, in the entire i97i year from January 1 to March 2010! In the historical time interval, the number of times the grouping pattern 11854 is the same as that of the March 1st of the year of the year is 1783. It should be noted that (4) is that the field 14 in the list 140a generated by the grouping mode in FIG. 8 is the same as that of all companies having the same five-day group arrangement. As described above, the same group number of positions 144a of the position 144a in the list generated by the clustering mode of the present invention may also be obtained from the data of the specific individual stock only. Next, as shown in step 35, Probability and Average Reward Calculation Module 14 Operation January 1, 1971 to February 2, March] Day Historical Time Zone The same permutation grouping pattern appears in the room, and after each day, the above is the rising probability (the block M5a) and the rising average compensation (settlement 142a), the falling probability of falling (block H6a) and the falling average reward (column) Bit 143a), and the flat disk probability (the flat disk probability field is not shown in Figure 8), wherein the rising probability, the lower 26 201222455 falling rate and the flat probability are the sum and the rising probability and falling The probability is available to the decision makers to decide whether to buy or sell the particular stock. Then, the standard investment stock selection system 1g is followed by the subsequent step 36·39, so that the list 148a, 149a# 15〇a in the list 1 of FIG. 8 and the portfolio list 16〇& in FIG. 9 can be obtained. Expected remuneration risk = 1 / unit risk expected remuneration. As shown in FIG. 9, the maximum risk of taking the risk per unit of expected remuneration is the function of the objective k. In the preferred embodiment, when the predetermined selection criterion is the expected remuneration (field 166a) is negative, the upper service rate (Field 165a) is less than the predetermined probability threshold of 0.5, the standard deviation _ _ secret) is less than 4, the recent month's revenue growth (field 162a) is less than 20%, and the 8th RSI (field 163 magic value is greater than 85) At this time, the decision and screening module 16 will generate a list 16 as shown in FIG. 9 such as the eighth block 169, which is the daily business group plus the first 4 days (ie 5 miles in total). The portfolio list engineering in this Figure 9 is sorted by the unit expected salary commitment risk field 167a and displayed on the stock picking system operation interface 23 of the display device 23 of the investor device 2, so that the investor can Bit 167a (or sorting field i61a) to select the stocks that are sold by March 2, 2, and 2, for example, to sell the stock of the company with the lowest unit expected return risk - 74.823. See Figure 4. 1 〇' In other embodiments of the present invention, the present invention adopts a combined grouping mode in FIG. After step 33 is completed, then CApM (Capital)

Asset Price Model) or the Securities Market Line (SML) is based on the theory of 'single index model' or arbitrage Pricing Therory 'APT' 〆*> 27 201222455 , establish a multi-factor model (Multifact〇r M〇dels). Empirically, the index model can be used to establish the predictive model of the regression model by using the index model and the reward model of the excess reward (relative to the risk-free compensation), or by adding other factors such as the government public debt excess compensation total factor. . A regression model of predictive simulation is established by the regression operation module 17, and the error is detected by the test method of the model prediction ability. Taking the general form of the Security Characteristic Une as an example, establish a regression equation (6):

Yt=f(Xt)+et (6) Yt is the daily remuneration of individual stocks; Xt is the weighted index 曰 remuneration. The stock type found by the investment stock selection method is used as the control variable &, then its general formula can be changed to the complex regression model as follows: «(xt; control variable) +et (7) or W stock selection method The type (different control variables) distinguishes the company's compensation data of different attributes' and then establishes several regression equations based on (8) for comparison. In addition, the test method for model prediction capability includes the following: (1) RMS error = V ^ f = l where >7 is a wide analog or predicted value; Γ is the actual value; τ is The number of simulation periods. (2) Root mean square error (RMS percent Err〇 = 201222455 (3) Mean simulation error (Mean Simulation e: 丄7^) (4) ^ ^ ^ tb ^ ^ (Mean Percent Error) = 1 ^ Yt

(5) Theil inequality coefficient, referred to as the υ value:

》'Fig. 1G Regression Results 17G is a regression analysis of the case of the conventional knife group by taking the Taimu stock as an example, and the method of the present invention is directed to the number of the group 1 in the five muds of the Taimu company and the group is the group respectively. Bu 3, 4, 5, 6

The historical reward data of the combined grouping model is subjected to regression analysis, and the regression result is calculated by using the == and other virtual variables to calculate the prediction error as the implementation. From the statistical point of view, The adjustment of the regression result of the mouth mud know model is lower than the adjusted R square value of 0.7749, which can prove that the invention uses the industry and the market as the standard to divide the stock market history data. For the eight major groups, the complex regression model is used to explain the variation of daily stock returns. The force is obviously better than the conventional situation that is not grouped by industry and the market. In the vehicle, the regression calculation module 17 can then calculate the relevant data of the model of the invention of the 1G regression result ,0, and calculate the predicted daily reward of the platform mud with the predicted weighted stock price. (4) FIG. 5' In other embodiments of the present invention, in FIG. 2, the present invention adopts the step 34 of the core group mode to complete the payment, and the same can be repeated as described above. 29 201222455 The model of returning to the historical reward data for grouping by arrangement pattern Calculate the predicted daily stock returns for a particular stock. In summary, the invention of the standard stock investment selection method is to conduct in-depth exploration of the primary information published by the stock exchange by the industry and the market for the market, and to use the results of the comparison between the two and the two to gradually classify Classify similar market performances as 'reuse of prior information. Subdivide the historical performance of stocks of the same category, gradually narrow the range of stock selection, and then use the relative number of times or Bayesian to calculate the future stocks. The probability of falling and the expected reimbursement value provide valuable and valuable information to investors, so that the purpose of the invention can be achieved. The above is only the preferred embodiment of the present invention, and the scope of the present invention is not limited thereto, that is, the simple equivalent change and modification of the patent scope and the description of the invention according to the present invention. All remain within the scope of the invention patent. BRIEF DESCRIPTION OF THE DRAWINGS FIG. 1 is a system architecture diagram illustrating a preferred embodiment of the standard stock investment stock selection method of the present invention, which is a standard, a stock, a database, and an investor device; Only Figure 2 is a flow chart illustrating the standard stock investment method of the present invention = group mode for grouping historical stocks of specific stocks: The method of grouping in the stock picking method is compared with Figure 3, which is a flow chart illustrating the present The invention uses the permutation grouping model to determine the historical remuneration data of the individual stocks; 30 201222455 Describes the historical remuneration data of the re-entry _ type of real stocks A preferred embodiment of grouping and relaying the daily stocks of the model 1; using the row=8 is the flow chart' to illustrate the standard stock investment method of the invention, and using the knife type to record the history of the specific stocks (four) A preferred embodiment for performing grouping and following a complex regression model to calculate individual stock returns;

The display is a schematic diagram of the operation interface, which is shown in the investor-side device. (4) The operation system of the stock system, which allows the stock selection system to operate the risk-based period 1 Embodiment; ^ 夕翻图. . 7 is an operation interface diagram 'Description of the stock selection system operation interface displayed on the investor device device', wherein the stock selection system operation interface 8 shows the investment power a list generated by the method of the present invention using the combined grouping mode. DETAILED DESCRIPTION OF THE PREFERRED EMBODIMENTS Figure 8 is a schematic diagram of an operation interface showing the operation of the stock selection system on the display device of the investor device, wherein the method of the stock selection system does not utilize the method of the present invention. An embodiment of the generated risk expectation reward list; circle 9 is an operation interface diagram illustrating the stock selection system operation interface displayed on the display of the investor device, wherein the stock selection system operation interface is '.·> A preferred embodiment of a portfolio list generated by the method of the present invention using an aligned clustering mode; and FIG. 10 is a schematic representation of a regression model illustrating a conventional regression model and a complex regression model in the method of the present invention for a platform mud The company's portfolio grouping model 31 201222455 type historical compensation data for regression analysis, the resulting regression results.

32 201222455

[Description of main component symbols] I ..........System-side server 10.........Standard investment stock selection system II .........Stock stock information The computing module 12 ... ... grouping module 13 ... ... grouping module 14 ... ... probability and average reward computing module 140... .... Risk Expected Remuneration List 14〇a... Risk Expected Remuneration List 1411-1412 Fields 141-149. Fields 142a-149a Shed 15 ......... Risk Expected Remuneration Operation module 150-150a Field 16......... Decision and screening module 160....... Portfolio list 160a... Only combination list 161 a-167a Block 161-169· Field 17......... Regression Module 170....... Regression Result 2 .......... Investor Side Set 21. ........ Processor 22......... Memory 220....... Select System System Interface Program 23.... ....... stock picking system operation interface 30-39 ... • step 40......... step 8 .........., stock price database 9 .... ... • Internet 33

Claims (1)

201222455 VII. Scope of application for patents: 1. - Standard method for investment stock selection 'for an investor to choose at least one suitable investment target' The standard investment stock selection method includes the following steps: (A) According to a stock price data The calculation of a historical remuneration data of all stocks in all the first-time intervals in the historical time interval of the most recent time interval, and the historical remuneration data of each industry classification; (B) according to the historical time interval Whether the remuneration of the industry in which all stocks in the first-time interval belong is greater than the historical remuneration of the market, whether the historical remuneration of individual stocks is greater than the historical remuneration of the stocks in which the stocks belong, and whether the historical remuneration of individual stocks is greater than the historical remuneration of the stocks, and all stocks in all - time intervals are grouped to generate groups of all stocks in all first time intervals; and (C) according to a particular grouping pattern of - the particular stocks in the group including the second = two of the most recent time interval, The same Μ group mode appears in the operation of the history and the following - the first time zone 4 The rising probability and the rising average (4) The same specific grouping mode occurs within the same period, and the subsequent specific grouping mode is the same as the first time-time and the first time is the sum of the probability, the falling probability and the flat probability. And the chance of a rise and the chance of a fall are the sale and purchase of that particular stock. The decision of W is 2. According to the method of stock selection and stock selection described in the scope of patent application ,1, in the process of grouping (B) in 34 201222455, the group is generated in eight steps, the specific group 楹4 people, ~ 6 mourning (c) The number of stocks in the range including the most recent time interval; the number of the eight groups - the specific group and the group in the recent time = the corresponding stocks in the comparison The time zone reward data is combined and grouped. 3_: According to the standard investment stock selection method described in item 1 of the patent application scope, in step 1 (c), the specific clustering mode is a permutation grouping mode, according to the specific time of the specific stock in the second time including the most recent time interval The group arrangement of the consecutive first time intervals in the district gate, the J stocks and all the stocks are selected - the historical reward data in the historical time interval is arranged and grouped. 4. According to the standard stock investment selection method described in item 1 of the patent application scope, 1 (10) further includes the group-specific grouping mode according to the specific individual stock in the second time interval: Out = the same group mode and then - (four) - (four) interval the specific increase in the number - the number of people, the number of declines in the fall, the number of times of the fall and the total number of times - the total number of times equals the number of times, the number of people The sum of the number of second and thousand times, and the probability of the upper service is equal to the total number of times. The amount of the decline is divided by the number of declines. 5. The target investment method of stock selection mentioned in item 1 of the patent scope is: in the step (C), the system uses the historical reward data and the degree function. Calculate the chance of the captain and the chance of falling. 35 201222455 According to the method of stock selection and stock selection described in the second paragraph of Shen β Patent I, in the “Hai (C) step, the conditional probability is used to calculate the rising machine and the falling probability. According to the U (4) item The standard investment selection method, wherein in the step (c), the Bayesian decision rule is used to calculate the rising probability and the falling probability. 8. According to the patent application scope, the standard investment selection The stock method, after the step (c), the standard stock picking method further comprises the following steps: (D) calculating one of the specific stocks expected salary = the rising probability x the rising average return + the falling probability χ the falling Average remuneration, (Ε) Operation of one of the specific stocks Unit risk expected return = the expected return divided by the risk, where the risk is calculated from the total risk, system risk and risk value One of the groups; '(F) for each stock, perform these (C) to (Ε) steps to obtain the probability of rising all units and the expected risk of unit risk; and (G) The predetermined screening criteria for screening reference data 'screen out the stocks of all the stocks that meet the predetermined screening criteria of the group, and sort the selected stocks by the unit risk expectation rewards from large to small to obtain a portfolio. a list for the investor to decide whether to buy or sell individual stocks in the portfolio list. 9. According to the standard stock selection method described in item 1 of the patent application scope, after the step (C), the label investment The stock picking method also includes the following steps: (H) Run nose one of the specific stocks expected salary = the probability of rising X on the 36 201222455 rising average return + the probability of falling x the average return of the fall; (1) computing one unit of the particular stock Risk Expected Remuneration = the expected remuneration divided by the risk 'and - the unit expected remuneration risk = 1 / the unit risk expected remuneration' where the risk is calculated from the total risk, system risk and risk expressed by standard deviation One of the groups; (7) for each share, carry out the steps (C), (H) and (1) to know the probability of rise of all stocks and The unit expects the remuneration to bear the risk; and (K) based on a set of predetermined screening criteria including a plurality of screening reference data, the division selects the stocks of all the stocks that meet the predetermined screening criteria of the group and treats the selected stocks as the unit's expected remuneration Large sorting to obtain a portfolio list for the investment two: Decide whether to buy or sell individual stocks in the portfolio list. 10. According to the standard stock investment stock selection method described in item 8 or 9 of the patent application scope, The screening reference data is selected from the group consisting of expected rewards, attendance, standard deviation values, fundamental data, price-based technical indicators, and quantity-based technical indicators. According to the standard stock investment stock selection method described in item i of the patent application scope, after the step (C), the standard stock investment stock selection method further comprises: using a complex regression model to calculate a daily stock return forecast value of the specific stock. . 12. A computer program product, wherein the program is loaded by a computer, and the method for selecting a stock selection method according to any one of the above-mentioned claims, wherein the computer-readable recording medium is There is recorded a code that can be controlled by a computer, and the program code includes the execution steps of the standard stock selection method as described in any one of the above-mentioned claims. 38