TWI799281B - Method and non-transient computer-readable recording medium for estimating portfolio-like efficiency allocation - Google Patents

Abstract

The present invention provides a method for estimating an investment portfolio by class-efficiency allocation, which uses executable codes to execute the steps of restriction condition setting, investment portfolio setting, generation of expected return data set, and class-efficiency allocation, mainly based on the investment of assets A collection of multiple investment portfolios is randomly generated when the amount meets a restriction condition, and the estimated annualized Sharpe value in the collection meets an estimation condition, and the estimation is used as a class efficiency allocator for use as The reference basis for selecting investment portfolios. The present invention also provides a non-transitory computer-readable recording medium capable of being read to perform the above method.

Description

Method and non-transient computer-readable recording medium for estimating portfolio-like efficiency allocation

The present invention relates to a method for estimating an investment portfolio, in particular to a method for estimating an investment portfolio with efficiency allocation, and a non-transitory computer-readable recording medium capable of executing the method.

The efficient allocation of conventional investment portfolios lies in that a portfolio contains multiple assets (such as stocks, bonds...), assuming that the sum of the investment proportions of these multiple assets is 1 (and does not use short-selling as an operation), through mean-difference optimization (Mean-Variance Optimization) Find the same expected rate of return and the least risky investment portfolio among multiple investment portfolios. Each investment portfolio found has the following two characteristics: 1. When the expected rate of return is fixed, Minimize risk (standard deviation). 2. With the risk (standard deviation) fixed, the expected rate of return can be maximized.

However, in the above-mentioned homogenous optimal efficiency allocation, the proportion of each asset obtained may not be an integer (that is, an odd number with a decimal point), which in practice causes the investment amount of the estimated investment portfolio to be inconsistent with the transaction amount. There are trading restrictions on incremental amounts, so it is not possible to directly put an efficient allocation of investment portfolios into the trading market for subscription.

In addition, due to the complexity of the calculation process of the homogeneous optimization, the software and hardware of the system must meet high specifications in order to perform smoothly.

Therefore, how to solve the above-mentioned problem of conventional efficiency allocation is the key point of the present invention.

In order to achieve the above purpose, the inventor has exhausted his mind and mind to study, and then developed a method for estimating investment portfolios of a kind of efficiency allocation, and a non-transient computer-readable recording medium, and the estimated investment amount can meet the requirements of the transaction. Incremental amount, so that the estimated investment portfolio can be directly put into the linked trading market for subscription.

The present invention provides a method for class efficiency allocation estimation investment portfolio, which is executed by an electronic device with a plurality of executable codes. The method includes constraint condition setting, investment portfolio setting, generation of expected return data, collection and class efficiency Configuration steps. First, select multiple assets to invest in, and set a restriction on each of the assets that meets the incremental unit of the amount; then set up an investment portfolio with some of the assets, the investment portfolio includes the investment amount of the several assets, and the investment of the several assets The amount is set to meet the restriction; and in the investment portfolio, the certain assets correspond to a historical return data retrieved from a database with individual investment amounts to generate an expected return data, and the certain assets are individually invested The investment weight is calculated based on the ratio of the amount to the total investment, and the expected return data includes at least an estimated annualized rate of return, an estimated annualized standard deviation, and an estimated annualized Sharpe value; repeated through an incremental amount change Execute the steps of setting the investment portfolio and generating the expected return data to generate a set including a plurality of the investment portfolios; finally, the investment portfolios whose estimated annualized Sharpe values in the set meet an estimation condition are estimated as class efficiency allocator.

The present invention also provides a non-transitory computer-readable recording medium, which stores a plurality of executable codes, so that an electronic device can execute the above method after reading and executing the executable codes.

In one embodiment, after the step of aggregating, a step of calculating the slope is further included, which is to sort the estimated annualized standard deviations of the multiple investment portfolios according to the estimated annualized standard deviations in accordance with The investment portfolio of a first-condition person is an initial benchmark point, and the estimated annualized rate of return and estimated annualized standard deviation of the remaining investment portfolios and the estimated annualized rate of return and estimated annualized rate of return corresponding to the initial benchmark point The slope calculated by normalizing the standard deviation, and then sorting each of the slopes, so as to obtain the investment portfolio whose slope meets a second condition from the plurality of investment portfolios, which is deemed to meet the estimation condition.

In one embodiment, after the step of calculating the slope, a step of loop operation is further included, which is to update the collection by removing those with slopes lower than the second condition among the plurality of investment portfolios, and use the slope to meet the The investment portfolio for the second condition is an updated reference point replacing the initial reference point, and then return to the step of calculating the slope to obtain the next investment portfolio whose slope meets the second condition.

In one embodiment, in the step of allocating efficiency, the initial reference point and at least one pitch group of slopes corresponding to the updated reference point are sorted from high to low. And the one with the highest estimated annualized Sharpe value is deemed to meet the estimation conditions. In one embodiment, in the step of configuring the type of efficiency, a type of efficiency front curve is drawn from the initial reference point and at least one of the updated reference points, and the estimated annualized The one with the highest Sharpe value is deemed to meet the estimation conditions.

In one embodiment, in the step of the loop operation, if the estimated annualized standard deviation and the estimated annualized rate of return are both lower than the slope and meet the second condition, it is excluded from the set of the investment portfolio.

In one embodiment, the investment amount is an upper limit of investment amount, a lower limit of investment amount, and an incremental amount of each asset in each investment portfolio.

In one embodiment, the upper limit of the investment amount is set by an input interface; the lower limit of the investment amount and the incremental amount are automatically brought out according to the subscription limit and transaction limit known to each asset.

In one embodiment, the historical return data is obtained by extracting a historical return rate corresponding to each of the assets from the database in the same time period when the listing time overlaps.

In one embodiment, the historical return data of each asset includes annualized rate of return and annualized standard deviation, and a correlation coefficient and covariate matrix between each asset.

In one embodiment, among the plurality of investment portfolios, the investment portfolio with the lowest estimated annualized standard deviation meets the first condition, and the investment portfolio with the highest slope meets the second condition By.

Thereby, the method for estimating an investment portfolio similar to efficiency allocation and the non-transitory computer-readable recording medium of the present invention can generate investment portfolio results similar to efficiency allocation, and since the investment amount conforms to the incremental amount at the time of transaction, the generated The investment portfolio can be directly put into the trading market for purchase, so as to achieve a more intuitive and efficient effect of the estimation result of the investment portfolio.

In order to fully understand the purpose, features and effects of the present invention, the present invention will be described in detail through the following specific embodiments and accompanying drawings, as follows:

The present invention provides a method 100 for estimating investment portfolios for efficiency allocation, which is executed by an electronic device (not shown in the figure) with a plurality of executable codes, and please refer to FIG. 1 , the method 100 includes setting constraints 101. The steps of investment portfolio setting 102, generation of expected return data 103, collection 104, calculation of slope 105, loop operation 106, and class efficiency allocation 107 are executed, wherein:

The step of setting the restriction conditions 101 is to select a plurality of assets to be invested, and to set a restriction condition for each of the assets that meets the increasing amount unit. The assets mentioned can be funds, stocks or bonds, but are not limited thereto. The incrementing unit of the amount mentioned above is the incrementing amount unit of the investment amount of the assets mentioned above, which can be directly put into the market for subscription without further conversion or processing (such as rounding or unconditional rounding down, etc.). In this embodiment, the multiple assets to be invested are selected from the three funds of Fund A, Fund B, and Fund C shown in Table 1, but the present invention is not limited to the three funds mentioned above. Table 1 (restriction setting) fund name Maximum investment amount Minimum investment amount Incremental amount Fund A 5000 1000 1000 Fund B 5000 1000 1000 Fund C 5000 1000 1000 Calculation data interval start date 2021/01/01 end date 2021/12/31 risk free rate 1.5%

In one embodiment, the investment amount is an upper limit of investment amount, a lower limit of investment amount, and an incremental amount of each asset in each investment portfolio. In one embodiment, the upper limit of the investment amount is set by an input interface; the lower limit of the investment amount and the incremental amount are based on the subscription restrictions and transaction restrictions known to each asset (for example, announced by a stock exchange company) limit) and automatically brought out. The upper limit of the investment amount is used to determine the maximum amount to be invested in the transaction for each asset in each of the generated investment portfolios (corresponding to the weight of each asset).

The step of setting investment portfolio 102 is to set up an investment portfolio with several assets, the investment portfolio includes the investment amount of the several assets, and the investment amount of each asset is set to meet the restriction condition. For example, the investment portfolio shown in Table 1 is a combination of three funds, including Fund A, Fund B, and Fund C. Both are 1,000 yuan, and the incremental amount is 1,000 yuan (thousand yuan) as the unit, and the difference in the incremental amount is used to generate a range that satisfies the lower limit of the investment amount and the upper limit of the investment amount, including Fund A, Fund B, and Multiple investment portfolios of Fund C. For example, in one investment portfolio, the investment amounts of Fund A, Fund B, and Fund C are respectively set at 1,000 yuan, 5,000 yuan, and 3,000 yuan. At this time, the total investment is 9,000 yuan, and so on (and See Group 1) in Table 5.

The step of generating expected return data 103 is that in the investment portfolio, the individual investment amounts of the assets correspond to a historical return data extracted from a database to generate an expected return data. The database can be Cloud server, or the local server of the electronic device.

In one embodiment, the expected return data is shown in Table 2, including an estimated annualized standard deviation, an estimated annualized Sharpe value, and an estimated annualized rate of return, wherein the estimated annualized Sharpe value It is calculated based on the corresponding estimated annualized standard deviation and estimated annualized rate of return. The calculation process takes into account the risk-free interest rate, which is preset at 1.5% in this example (calculation formula: Sharpe value = (rate of return - risk-free rate)/standard deviation). In this step, the investment weight is calculated based on the ratio of the individual investment amount of the certain assets to the total investment.

Following the previous example, the total investment of Fund A, Fund B, and Fund C is 9,000 yuan, and the investment weights of Fund A, Fund B, and Fund C can be calculated as 11.11%, 55.56%, and 33.33%. The calculated estimated annual The standard deviation is 7.323, the estimated annualized rate of return is 11.781, and the estimated annualized Sharpe value is 1.404 (see also Group 1 in Table 2). Table 2 (Expected Compensation Information) Fund A Fund B Fund C Estimated Annualized Standard Deviation Estimated Annualized Rate of Return Estimated Annualized Sharpe Value group Investment weight 1 11.11% 55.56% 33.33% 7.323 11.781 1.404 2 22.22% 33.33% 44.44% 5.013 8.362 1.369 3 50.00% 16.67% 33.33% 6.145 6.133 0.754 4 20.00% 40.00% 40.00% 5.739 9.407 1.378 5 45.45% 27.27% 27.27% 7.013 7.778 0.895 6 33.33% 11.11% 55.56% 3.528 4.943 0.976 7 23.08% 38.46% 38.46% 5.846 9.210 1.319 8 33.33% 33.33% 33.33% 6.297 8.553 1.120 9 71.43% 14.29% 14.29% 9.218 6.115 0.501 10 33.33% 44.44% 22.22% 8.134 10.359 1.089 11 9.09% 45.45% 45.45% 5.461 10.105 1.576 12 14.29% 14.29% 71.43% 2.266 ( minimum standard deviation) 5.130 1.602 13 71.43% 14.29% 14.29% 9.218 6.115 0.501 14 45.45% 45.45% 9.09% 10.190 10.732 0.906 15 45.45% 9.09% 45.45% 4.645 4.823 0.715 16 14.29% 71.43% 14.29% 10.875 14.415 1.188 17 44.44% 22.22% 33.33% 6.137 6.939 0.886 18 55.56% 33.33% 11.11% 9.614 8.936 0.773 19 33.33% 50.00% 16.67% 9.166 11.262 1.065 20 12.50% 25.00% 62.50% 3.137 6.840 1.702

In one embodiment, the historical return data is obtained by extracting a historical return rate corresponding to each of the assets from the database in the same time period when the listing time overlaps. As shown in Table 1, the same time period in which the listing time overlaps is based on 2021/1/1~2021/12/31 as the data calculation interval, but it is not limited to the time period in this example. The same time period Can be longer or shorter as long as time to market overlaps. As shown in Table 3, the historical return data includes historical daily return rates of Fund A, Fund B and Fund C, but the present invention is not limited thereto. Table 3 (historical daily rate of return) date Fund A Fund B Fund C 2021/1/1 0 0 0 1/4/2021 -0.106846 -1.663181 -0.26647 2021/1/5 1.215025 0.456926 -0.161113 2021/1/6 1.09799 -0.280763 0.122431 2021/1/7 0.936944 1.698374 0.109785 2021/1/8 0.854009 0.945725 0.106633 2021/1/11 -1.084632 -0.652207 -0.171646 2021/1/12 1.000864 -0.314253 -0.075836 2021/1/13 -0.548619 0.044175 0.012986 2021/1/14 0.851837 -0.624576 0.234335 2021/1/15 -1.679984 -0.487478 0.027154 2021/1/18 0.016032 0.016032 0.016032 2021/1/19 0.854172 1.03977 0.037539 2021/1/20 0.811446 1.640245 0.16065 (omitted) 2021/12/27 0.591084 1.243687 -0.106496 2021/12/28 -0.095572 -0.484715 -0.095572 2021/12/29 -0.309873 0.093317 0.151419 2021/12/30 0.010989 -0.101554 0.041611 2021/12/31 0.138944 -0.291721 -0.000138

As shown in Table 4, the historical return data of Fund A, Fund B, and Fund C are based on individual historical daily return rates (as shown in Table 3), and further include rate of return, standard deviation, annualized rate of return and annualized rate of return. standard deviation. Table 4 (Historical Compensation Data) fund name Fund A Fund B Fund C rate of return 0.0170 0.0746 0.0102 standard deviation 0.8944 1.0007 0.2404 Annualized rate of return 4.2859 18.8117 2.5627 annualized standard deviation 14.1978 15.8862 3.8158

The step of randomly generating a set 104 is to repeatedly execute the step of setting the investment portfolio through incremental changes in the amount, and randomly generate a set including a plurality of the investment portfolios. In one embodiment, the random generation of the plurality of investment combinations is performed by using an application program on a computer, but the present invention is not limited thereto. As shown in Table 5, a collection of 20 investment portfolios including groups 1-20 is randomly generated, and Fund A, Fund B, and Fund C of each investment portfolio are all in line with the aforementioned upper limit of investment amount (5,000 yuan) and investment amount The lower limit (1,000 yuan), and the condition that the incremental amount is in units of 1,000 yuan. Table 5 (collection of portfolios) Fund A Fund B Fund C Total investment group investment amount 1 1000 5000 3000 9000 2 2000 3000 4000 9000 3 3000 1000 2000 6000 4 2000 4000 4000 10000 5 5000 3000 3000 11000 6 3000 1000 5000 9000 7 3000 5000 5000 13000 8 4000 4000 4000 12000 9 5000 1000 1000 7000 10 3000 4000 2000 9000 11 1000 5000 5000 11000 12 1000 1000 5000 7000 13 5000 1000 1000 7000 14 5000 5000 1000 11000 15 5000 1000 5000 11000 16 1000 5000 1000 7000 17 4000 2000 3000 9000 18 5000 3000 1000 9000 19 2000 3000 1000 6000 20 1000 2000 5000 8000

The step of calculating the slope 105 is to sort the estimated annualized standard deviations of the multiple investment portfolios, and use the investment portfolio whose estimated annualized standard deviation meets a first condition as an initial reference point (marked as (X1, Y1) in the coordinates in Figure 2). The investment portfolio that meets the first condition, in one embodiment, may be the investment portfolio with the lowest estimated annualized standard deviation among the multiple investment portfolios, such as the investment portfolio of Group 12 in Table 2 , and its "estimated annualized standard deviation" is 2.266, which is the lowest in the set.

Carrying on from the above, and then based on the estimated annualized rate of return and estimated annualized standard deviation of the rest of the investment portfolio, the slope calculated from the estimated annualized rate of return and estimated annualized standard deviation corresponding to the initial benchmark point, and then Sorting the magnitude of each of the slopes, so as to obtain the highest slope from the plurality of investment portfolios, which meets a second condition, for example, in the coordinates of Figure 2, the investment portfolio marked as (X2, Y2) belongs to it. Group 12 has the highest slope of 1.963 (see Table 6 for the current estimated annualized standard deviation of 2.266, the current estimated annualized rate of return of 5.130, and the current group's slope of 1.963).

The step of the loop operation 106 is to use the investment portfolio with a slope higher than the second condition as an updated reference point among the plurality of investment portfolios (in this way, the updated reference point replaces the initial reference point as a new reference point ), and delete the one with the lower rate of return to update the set, and then return to the step of calculating the slope to obtain the next investment portfolio condition with the highest slope. The person who meets the second condition, in addition to the investment portfolio with the highest slope mentioned above, may further be the person whose estimated annualized standard deviation is lower than the slope and meets the second condition, and the estimated annualized return The one whose rate of return is higher than the slope and meets the second condition is the one with the lower rate of return, so as to be removed from the set. The "current" refers to the initial benchmark point or the group corresponding to the updated benchmark point, the estimated annualized standard deviation, and the estimated annualized rate of return as the basis of comparison.

For example, as shown in Table 6, according to the investment portfolios of groups 1-20, the steps of the loop operation 106 are performed five times in total, and the results are represented by point numbers 1-5 of the leading edge in Table 6 respectively. The point number 1 of the leading edge line is based on the minimum estimated annualized standard deviation of group 12 as the initial reference point, and then compares the slope with the other groups, among which group 20 has the largest slope of 1.963. Next, among groups 1 to 19, those whose estimated annualized standard deviation and estimated annualized rate of return are both lower than group 20 (i.e. group 6, group 12, and group 15 are excluded), Change to group 20 with the largest slope as the first update reference point, and then compare the slope with the remaining groups after elimination, and so on, to obtain the second update with group 11 having the largest slope The reference point, the third updated reference point with group 1 having the largest slope, and the fourth updated reference point with group 16 having the largest slope, so as to obtain group 12, group 20, group Point numbers 1~5 of the five leading edge lines including category 11, group 1, and group 16. Table 6 (Maximum slope and group of loop operation) Point number of leading edge line 1 2 3 4 5 current group 12 20 11 1 16 Current Estimated Annualized Standard Deviation 2.266 3.137 5.461 7.323 10.875 Current Estimated Annualized Rate of Return 5.130 6.840 10.105 11.781 14.415 group Slope with current group 1.315 1.180 0.900 0.000 0.000 1 1.177 0.811 0.000 0.000 0.000 2 0.258 0.000 0.000 0.000 0.000 3 1.232 0.987 0.000 0.000 0.000 4 0.558 0.242 0.000 0.000 0.000 5 0.000 0.000 0.000 0.000 0.000 6 1.140 0.875 0.000 0.000 0.000 7 0.849 0.542 0.000 0.000 0.000 8 0.142 0.000 0.000 0.000 0.000 9 0.891 0.704 0.095 0.000 0.000 10 1.557 1.405 0.000 0.000 0.000 11 0.000 0.000 0.000 0.000 0.000 12 0.142 0.000 0.000 0.000 0.000 13 0.707 0.552 0.133 0.000 0.000 14 0.000 0.000 0.000 0.000 0.000 15 1.079 0.979 0.796 0.742 0.000 16 0.467 0.033 0.000 0.000 0.000 17 0.518 0.324 0.000 0.000 0.000 18 0.889 0.733 0.312 0.000 0.000 19 1.963 0.000 0.000 0.000 0.000 20 maximum slope 1.963 1.405 0.900 0.742 Maximum slope group 20 11 1 16

The step of the quasi-efficiency allocation 107 is estimating the portfolios that meet an estimation condition with the estimated annualized Sharpe value in the set as quasi-efficiency allocators. In one embodiment, in the step of this type of efficiency configuration, the slopes corresponding to the initial reference point and at least one update reference point are sorted from high to low, and the one with the highest estimated annualized Sharpe value is regarded as to meet the estimation criteria. According to the initial benchmark point and the four updated benchmark points, the corresponding group/estimated annualized Sharpe values shown in Table 7 are Group 20/1.702, Group 12/1.602, Group 11/1.576, Class 1/1.404, and Class 16/1.188.

In one embodiment, in the step of this type of efficiency configuration 107, the initial reference point and four update reference points corresponding to the aforementioned group 1, group 11, group 12, group 16, and group 20, Based on this, a type of efficiency frontier curve (as shown in Figure 3) is drawn, and the one with the highest annualized Sharpe value in this type of efficiency frontier curve, in this example, is the investment portfolio of group 20. to meet the estimated conditions. Table 7 (data corresponding to each group point in the class efficiency frontier curve) group Estimated Annualized Standard Deviation Estimated Annualized Rate of Return Estimated Annualized Sharpe Value 12 2.266 5.130 1.602 20 3.137 6.840 1.702 11 5.461 10.105 1.576 1 7.323 11.781 1.404 16 10.875 14.415 1.188 risk-free rate point 0 1.5

The present invention also provides a non-transitory computer-readable recording medium, which stores a plurality of executable codes, so that the electronic device can execute the above method after reading and executing the executable codes.

The algorithm of the above-mentioned embodiment is listed and described as follows with its virtual code (Pseudo Code) here: " begin //Start Initialize FundsSet.InvAmt; //Set the upper limit, lower limit, and incremental amount of each fund in the investment portfolio Initialize FundsSet.HisReturn; //Get the historical rate of return data of each fund in the investment portfolio Function calFundsExpectReturnRisk(FundsSet ){ //Get the same or longer/shorter time period of the historical rate of return data in the above funds GetCommonHisReturnLength(FundsSet.HisReturn); //Calculate the estimated annualized rate of return and estimated annualized standard deviation of each fund CalExpectReturn (FundsSet); CalRisk(FundsSet); } for I = 0 & I < N;//Loop operation N times RandmAmtSet = RandomAmtAllocation(FundsSet); //Obtain a set of investment amount allocation randomly within the upper limit and lower limit of the investment amount of each fund AllocationInPercentSet = AllocationInPercent(RandmAmtSet); //According to the allocation amount of each fund above, calculate the total investment portfolio, and the individual weights of each fund in the investment portfolio Portolio.ExpectReturnRisk = CalPortfolio ExpectReturnRisk(AllocationInPercentSet);// Calculate the estimated annualized return of the investment portfolio according to the weight of each fund Rate and estimated annualized standard deviation PortfolioSet = AddPortofolio(Portolio);//Add to the set of investment portfolio end forLowestRisk = SortPortfolioRisk(PortfolioSet);//Sort the estimated annualized standard deviation of each portfolio and get the smallest one as the initial Benchmark; CalPortfolioSharpRatio; (PortfolioSet);//Calculate the estimated annualized Sharpe value of all portfolios SortPortfolioSharpRatio; (PortfolioSet);//Get the estimated annualized Sharpe value by sorting and get the largest point IniPoint = LowestRisk;/ / Portfolio with the smallest estimated annualized standard deviation in the portfolio as the initial benchmark point N = PortfolioSet; // Number of portfolios included for I = 0 & I < N; // Loop operation N times HighSlopPortfolio = CalHighSlop(IniPoint, PortfolioSet); //Get the points of all other portfolios whose estimated annualized rate of return and estimated annualized standard deviation are greater than the initial benchmark point, and calculate the slope of these points, and sort the slope values to obtain the slope The largest portfolio IniPoint = HighSlopPortfolio;//The point with the largest slope above is the update reference point of the portfolio PortfolioSet = PickHigherSlopExpectReturnLowerRisk(PortfolioSet); //Exclude all other low slopes, and the estimated annualized standard deviation and estimated year All portfolio points with lower rate of return, create a new portfolio collection The values of standard deviation and estimated annualized rate of return are sorted from small to large DrawQuasiEfficientFrontier(PortfolioSet);//Draw QuasiEfficientFrontier(PortfolioSet);//Draw the quasi-efficiency frontier curve FinalPortfolio = SortPortfolioSharpRatio (PortfolioSet); //Get the investment with the highest estimated annualized Sharpe value The combination becomes the optimized efficiency configuration end //End"

From the above description, it is not difficult to find that the characteristics of the present invention lie in the method of the present invention for quasi-efficiency allocation and estimation of the investment portfolio, and the non-transient computer-readable recording medium. Through the setting of the investment portfolio, the investment amount of the assets must meet the The increment unit of the amount is the restriction condition, and then refer to the historical return data to generate the expected return data as the estimation basis, and through repeated execution, in the set of randomly generated investment portfolios, the annualized Sharpe value is estimated to meet the estimation conditions. The combination is a quasi-efficient allocator, thereby producing a portfolio result that is close to efficient allocation, and since the investment amount conforms to the incremental amount at the time of the transaction, the resulting portfolio can be directly put into the trading market for subscription, so as to achieve the investment portfolio Estimation results are more intuitive and more efficient.

The present invention has been disclosed above with preferred embodiments, but those skilled in the art should understand that the embodiments are only used to describe the present invention, and should not be construed as limiting the scope of the present invention. It should be noted that all changes and substitutions equivalent to the embodiment should be included in the scope of the present invention. Therefore, the scope of protection of the present invention should be defined by the scope of the patent application.

100: method 101: Restriction setting 102:Portfolio Setup 103: Generate expected remuneration data 104: collection 105: Calculate the slope 106: Loop operation 107:Class Efficiency Configuration

Fig. 1 is a method flowchart of a specific embodiment of the present invention. Fig. 2 is a coordinate diagram of an investment portfolio based on an initial reference point according to a specific embodiment of the present invention. Fig. 3 is a graph of quasi-efficiency frontiers of a specific embodiment of the present invention.

100: method

101: Restriction setting

102:Portfolio Setup

103: Generate expected remuneration data

104: collection

105: Calculate the slope

106: Loop operation

107:Class Efficiency Configuration

Claims (12)

A method for estimating investment portfolio efficiency allocation, which is executed by an electronic device with a plurality of executable codes, the method includes: Restriction setting: select multiple assets to invest in, and set a restriction that meets the amount increment unit for each asset; Investment portfolio setting: setting up an investment portfolio with certain assets, the investment portfolio includes the investment amount of the certain assets, and the investment amount of the certain assets is set to meet the restriction; Generating expected return data: In the investment portfolio, the investment weight is calculated based on the ratio of the individual investment amount of the certain assets to the total investment, and the individual investment amount of the certain assets corresponds to a historical return data extracted from a database To generate an expected return data, and calculate the investment weight based on the ratio of the individual investment amount of the certain assets to the total investment, the expected return data includes at least an estimated annualized rate of return, an estimated annualized standard deviation, and an estimated estimated annualized Sharpe value; Aggregation: Repeating the steps of setting the investment portfolio and generating the expected return data by changing an incremental amount, and randomly generating a set including a plurality of the investment portfolios; and Class-efficiency allocation: The investment portfolio whose estimated annualized Sharpe value in this set meets an estimation condition is estimated as a class-efficiency allocator. The method for estimating investment portfolios for efficiency allocation as described in Claim 1, wherein after the step of aggregating, further includes a step of calculating the slope, which is sorted according to the estimated annualized standard deviations of the multiple investment portfolios , taking the investment portfolio whose estimated annualized standard deviation meets a first condition as an initial benchmark point, and according to the estimated annualized rate of return and estimated annualized standard deviation of the remaining investment portfolios corresponding to the initial benchmark point Calculate the slope of the estimated annualized rate of return and the estimated annualized standard deviation, and then sort the slopes in order to obtain a portfolio whose slope meets a second condition from the multiple investment portfolios, which is regarded as meet the estimation criteria. The method for estimating investment portfolios for efficiency allocation as described in claim 2, wherein after the step of calculating the slope, a step of circle calculation is further included, which is to eliminate the returns that are lower than those that meet the multiple investment portfolios For the slope of the second condition, update the set, and use the portfolio whose slope meets the second condition as an updated reference point to replace the initial reference point, and then return to the step of calculating the slope to obtain the next slope meeting the Portfolio for the second condition. The method for estimating investment portfolios for efficiency allocation as described in claim item 3, wherein, in the step of this type of efficiency allocation, the initial benchmark point and at least one investment group corresponding to the slope corresponding to the updated benchmark point are adjusted for high and low Sort, and the one with the highest estimated annualized Sharpe value is considered to meet the estimation conditions. The method for estimating investment portfolios of efficiency allocation as described in claim item 4, wherein, in the step of allocation of such efficiency, a type of efficiency frontier curve is drawn from the initial reference point and at least one update reference point, and the The one with the highest estimated annualized Sharpe value in the class-efficiency frontier curve is deemed to meet the estimation condition. The method for estimating investment portfolios for efficient allocation as described in claim item 5, wherein, in the step of the loop operation, the estimated annualized standard deviation and the estimated annualized rate of return are both lower than the slope Those that meet the second condition are investment portfolios excluded from the collection. The method for estimating investment portfolios for efficiency allocation as described in Claim 4, wherein the investment amount is an upper limit of investment amount, a lower limit of investment amount, and the incremental amount of each asset in each investment portfolio. The method for estimating investment portfolios for efficiency allocation as described in claim item 7, wherein the upper limit of the investment amount is set by an input interface; the lower limit of the investment amount and the incremental amount are known according to each asset The subscription limit and transaction limit are automatically brought out. The method for estimating investment portfolios for efficiency allocation as described in claim item 4, wherein the historical return data is extracted from the database in the same time period when the listing time overlaps, and a historical rate of return corresponding to each asset is individually extracted acquired. As described in claim item 9, the method for estimating investment portfolios for efficiency allocation, wherein, the historical return data to which each asset belongs includes annualized rate of return and annualized standard deviation, and a correlation coefficient between each asset and covariate matrix. The method of efficiency allocation estimation investment portfolio as described in claim item 10, wherein, among the plurality of investment portfolios, the investment portfolio with the lowest estimated annualized standard deviation is the one that meets the first condition, and The investment portfolio with the highest slope is the one meeting the second condition. A non-transitory computer-readable recording medium, which stores a plurality of executable codes, so that an electronic device can perform the following steps after reading and executing the executable codes, including: Restriction setting: select multiple assets to invest in, and set a restriction that meets the amount increment unit for each asset; Investment portfolio setting: setting up an investment portfolio with certain assets, the investment portfolio includes the investment amount of the certain assets, and the investment amount of the certain assets is set to meet the restriction; Generating expected return data: In the investment portfolio, the investment weight is calculated based on the ratio of the individual investment amount of the certain assets to the total investment, and the individual investment amount of the certain assets corresponds to a historical return data extracted from a database To generate an expected return data, and calculate the investment weight based on the ratio of the individual investment amount of the certain assets to the total investment, the expected return data includes at least an estimated annualized rate of return, an estimated annualized standard deviation, and an estimated estimated annualized Sharpe value; Aggregation: Repeating the steps of setting the investment portfolio and generating the expected return data by changing an incremental amount, and randomly generating a set including a plurality of the investment portfolios; and Class-efficiency allocation: The investment portfolio whose estimated annualized Sharpe value in this set meets an estimation condition is estimated as a class-efficiency allocator.

Images