KR102314793B1 - Method of providing asst management service and server performing the same - Google Patents

Abstract

The asset management server according to an embodiment uses a screening unit that selects candidate funds for each of a plurality of major asset groups through screening of funds sold by a financial institution, and a plurality of different strategies to select the candidate funds. A portfolio generation unit for each strategy that generates a plurality of portfolios corresponding to the plurality of strategies from A plurality of fund combinations are constructed from a plurality of funds, and a fund combination and asset weight having a high correlation coefficient with the result of synthesizing the plurality of portfolios according to the optimal weight among the plurality of fund combinations through a genetic algorithm are calculated as the final fund It may include an asset management server including a strategy synthesizing unit that determines the combination and the final asset proportion, and generates a final portfolio of the asset allocation model according to the final fund combination and the final asset proportion.

Description

A method of providing an asset management service and a server performing the same

The present disclosure relates to a method for providing an asset management service and a server for performing the same.

Individuals diversify their investments in multiple assets rather than investing in just one asset for risk diversification and maximum return. In other words, individuals try to obtain the best risk-return relationship by constructing a portfolio with a combination of several financial assets.

However, for this purpose, an individual can use the system such as accounting standards, taxation and various regulatory provisions, actual asset management, various market indices, such as price, interest rate, stock index, real estate price index, exchange rate, crude oil price, economic growth rate, trade balance, consumer There is a difficulty in acquiring information such as attitude index and unemployment rate and information on the comprehensive financial taxation system.

Accordingly, recently, the demand for individualized asset management services for financial institutions is gradually increasing.

Robo-advisor is an online asset management service that performs portfolio management through a mobile device or PC instead of a human private banker (PB) through advanced algorithms and big data. For asset allocation products operated with quantitative algorithms, such as robo-advisors, rules for determining the target and size of trading are defined in advance. These rules often have a problem that they cannot respond flexibly to the market as a single strategy in consideration of various financial environments.

Embodiments provide a method of providing an asset management service capable of providing asset allocation suitable for a market environment by flexibly responding to various financial environments, and a server performing the same.

In order to achieve the above or other object, the asset management server according to an embodiment includes a screening unit that selects candidate funds for each major asset class through screening of funds sold by a financial institution, a plurality of different A portfolio generating unit for each strategy that generates a plurality of portfolios corresponding to the plurality of strategies from the candidate funds using the strategy, and an optimal weight between the plurality of strategies based on the returns and volatility of the plurality of portfolios and constructing a plurality of fund combinations from a plurality of funds constituting the plurality of portfolios, and through a genetic algorithm, a correlation coefficient with the result of synthesizing the plurality of portfolios according to the optimal weight among the plurality of fund combinations and a strategy synthesizing unit for determining a high fund combination and asset ratio as a final fund combination and final asset ratio, and generating a final portfolio of an asset allocation model according to the final fund combination and the final asset ratio.

The strategy synthesis unit is configured to combine all the funds constituting the plurality of portfolios according to the asset weight given by the combination of the optimal weight and its weight allocated to each fund in each of the plurality of portfolios to synthesize the plurality of portfolios. can

The strategy synthesizing unit determines the asset weights of each of the plurality of fund combinations through the genetic algorithm, and the asset weighting with the highest correlation coefficient with the result of synthesizing the plurality of portfolios among the asset weights of each of the plurality of fund combinations may be determined as the final asset ratio.

The correlation coefficient may correspond to a degree of similarity between reference values.

Each of the plurality of portfolios and the final portfolio may include recommended funds selected for each of a plurality of major asset groups and asset weights of the recommended funds.

The screening unit may include at least one fund determined to be an overheated fund based on a relative strength index among the funds sold by the financial institution and a fund with a low return rate of less than or equal to a predetermined value for one month from among the remaining funds. , at least one fund selected in the order of high returns for each of the plurality of major asset groups may be selected as the candidate fund.

The plurality of strategies may include at least two of a momentum strategy, a Black Literman strategy, and a risk parity strategy.

The asset management server further comprises a service policy database for storing a service policy in which a risk limit and a weight limit for each asset type are defined for each investment type, and the portfolio creation unit for each strategy is configured to generate the plurality of portfolios based on the service policy. The risk rating and asset type of the constituent funds may be limited.

The optimal weight may be calculated by a difference between a value obtained by multiplying an average return for a predetermined period by a risk preference and volatility during the predetermined period.

In addition, the method of providing an asset management service of an asset management server according to an embodiment includes the steps of selecting candidate funds for each major asset class through screening of funds sold by financial institutions, and performing a plurality of different strategies. determining, from the candidate funds, a plurality of portfolios corresponding to the plurality of strategies, determining an optimal weight between the plurality of strategies based on returns and volatility of the plurality of portfolios, via a genetic algorithm. , from among a plurality of fund combinations combined from a plurality of funds constituting the plurality of portfolios, a fund combination and asset ratio having a high correlation coefficient with the result of synthesizing the plurality of portfolios according to the optimal ratio, the final fund combination and the final Determining the asset weighting, generating a final portfolio of an asset allocation model according to the final fund combination and the final asset weighting, and providing the final portfolio to a customer.

In the method, synthesizing the plurality of portfolios by combining all the funds constituting the plurality of portfolios according to the asset weight given by the combination of the optimal weight and the own weight allocated to each fund in each of the plurality of portfolios may include more

The determining of the final fund combination and the final asset proportion may include determining asset proportions of each of the plurality of fund combinations through the genetic algorithm, and the plurality of portfolios among the asset proportions of each of the plurality of fund combinations. and determining an asset proportion having the highest correlation coefficient with the result of synthesizing as the final asset proportion.

In the above method, the correlation coefficient may correspond to a degree of similarity between reference values.

In the method, each of the plurality of portfolios and the final portfolio may include recommended funds selected for each of the plurality of major asset groups and asset weights of the recommended funds.

The step of selecting the candidate funds may include selecting at least one fund determined to be an overheated fund based on a relative strength index among the funds sold by the financial institution and a fund having a low rate of return below a predetermined value for one month. The method may include selecting, as the candidate fund, at least one fund selected in the order of highest return for each of the plurality of major asset groups, among the remaining funds.

In the method, the plurality of strategies may include at least two of a momentum strategy, a Black Literman strategy, and a risk parity strategy.

The step of determining the plurality of portfolios may include limiting the risk class and asset type of the funds constituting the plurality of portfolios based on a service policy in which a risk limit and a weight limit for each asset type are defined for each investment type. can

The calculating of the optimal weight may include calculating the optimal weight based on a difference between a value obtained by multiplying an average return for a predetermined period by risk preference and volatility during the predetermined period.

According to embodiments, it is possible to flexibly respond to various financial environments and provide asset allocation suitable for market environments.

1 schematically illustrates an asset management service system according to an embodiment.
2 illustrates an example of selecting a fund combination included in a final portfolio through a genetic algorithm.
3 shows an asset allocation model theoretically synthesized by a plurality of portfolios for each strategy, and a performance trend and correlation coefficient for each investment propensity of an asset allocation model that is finally provided to a customer by the asset management server according to the embodiment.
4 schematically illustrates a method of providing an asset management service according to an embodiment.
5A and 5B are diagrams for explaining an effect of a method for providing an asset management service according to an embodiment.

Hereinafter, the embodiments disclosed in the present specification will be described in detail with reference to the accompanying drawings, but identical or similar components are given the same and similar reference numerals, and overlapping descriptions thereof will be omitted. The suffixes "module" and "part" for the components used in the following description are given or mixed in consideration of only the ease of writing the specification, and do not have a meaning or role distinct from each other by themselves. In addition, in describing the embodiments disclosed in the present specification, if it is determined that detailed descriptions of related known technologies may obscure the gist of the embodiments disclosed in the present specification, the detailed description thereof will be omitted. In addition, the accompanying drawings are only for easy understanding of the embodiments disclosed in the present specification, and the technical spirit disclosed herein is not limited by the accompanying drawings, and all changes included in the spirit and scope of the present invention , should be understood to include equivalents or substitutes.

Terms including an ordinal number such as first, second, etc. may be used to describe various elements, but the elements are not limited by the terms. The terms are used only for the purpose of distinguishing one component from another.

When a component is referred to as being “connected” or “connected” to another component, it is understood that the other component may be directly connected or connected to the other component, but other components may exist in between. it should be On the other hand, when it is mentioned that a certain element is "directly connected" or "directly connected" to another element, it should be understood that the other element does not exist in the middle.

The singular expression includes the plural expression unless the context clearly dictates otherwise.

In the present application, terms such as "comprises" or "have" are intended to designate that a feature, number, step, operation, component, part, or combination thereof described in the specification exists, but one or more other features It should be understood that this does not preclude the existence or addition of numbers, steps, operations, components, parts, or combinations thereof.

Hereinafter, a method of providing an asset management service according to an embodiment and a server performing the same will be described in detail with reference to the necessary drawings.

1 schematically illustrates an asset management service system according to an embodiment.

Referring to FIG. 1 , an asset management system according to an embodiment includes a database management system (DBMS) 100 , an asset management server 200 , a manager terminal 300 , a window terminal 400 , and a client terminal. (500).

The DBMS 100 may store and manage information for providing an asset management service into a database. The DBMS 100 may include a financial product database (DB) 110 , a service policy DB 120 , and a customer information DB 130 .

The financial product DB 110 may store and manage a list of assets (eg, funds) that can be sold by a financial institution having an asset management service system, and asset information of each asset. Here, the asset information may include information on the investment type of each asset (domestic stock type, overseas advanced stock type, emerging stock type, domestic bond type, overseas bond type, etc.), product name, yield, risk grade, and the like.

The service policy DB 120 may store and manage the service policy of the asset management service system. The service policy defines the restrictions when composing the portfolio of the asset allocation model.

The service policy may include information defining risk limits (minimum risk grade and maximum risk grade) for each investment propensity of the portfolio.

To this end, all funds sold in the asset management service system have a risk grade defined, and based on this, a risk limit can be defined for each investment propensity.

Table 1 below shows an example of classifying funds by risk class.

Fund risk rating Justice Grade 1 very high risk 2nd grade high risk 3rd grade somewhat high risk 4th grade moderate risk 5th grade low risk 6th grade very low risk

In Table 1 above, funds are classified into risk grades from grade 1 to grade 6, and the higher the grade, the lower the risk. The risk grade of each fund determined based on this may be stored as asset information of each asset in the financial product DB 110 . On the other hand, when the fund risk grade defined as shown in Table 1 is converted to the risk grade in the strategy for portfolio creation, as shown in Table 2 below, the risk level of the fund may be modified so that the higher the grade, the higher the risk.

Fund risk rating Risk Rating in Strategy Grade 1 6 2nd grade 5 3rd grade 4 4th grade 3 5th grade 2 6th grade One

And, in the service policy, based on the risk grade in the strategy defined in Table 2, the risk limit (minimum, maximum) for each investment propensity of the portfolio may be defined as shown in Table 3 below.

Customer Investment Propensity Minimum risk level hazard level max offensive investment 2 4 active investment 2 4 risk neutral 1.5 3 stability-seeking type One 2.5

The risk level for each investment propensity of the portfolio corresponds to the value calculated by applying the weight of each fund to the risk classes of a plurality of funds constituting the portfolio as a weight and then calculating the average. The scope is defined by investment propensity. For example, if a customer's investment propensity is an aggressive investment type, the risk grade of the portfolio may be limited to a minimum of grade 2 and a maximum of grade 4. The service policy may include information defining the weight limit for each asset type included in the portfolio for each investment propensity.

Table 4 below shows an example in which the weight limit for each asset type included in the portfolio is set for each investment propensity.

stock
Specific gravity minimum bond
Specific gravity minimum danger
Asset Share
Ieast danger
Asset Share
maximum 2nd grade
more
maximum
importance domestic
stock up Overseas
stock up Emerging stocks
maximum Individual
Fund Minimum
importance Individual
fund max
importance attack
investment type 50% 25% 25% 65% 30% 60% 50% 25% 10% 70% active investment 40% 35% 20% 55% 25% 50% 40% 15% 10% 70% risk neutral 20% 50% 0% 45% 10% 40% 30% 10% 10% 70% stable (seeking) type 10% 60% 0% 40% 0% 35% 30% 10% 10% 70%

Taking the portfolio provided to aggressive investment customers in Table 4 above as an example, among the assets that make up the portfolio, stocks make up at least 50%, bonds at least 25%, risky assets at least 25%, and up to 65 %, the proportion of assets with risk grade 2 or higher is up to 30%, domestic stocks up to 60%, foreign stocks up to 50%, emerging stocks up to 25%, and individual funds’ minimum weighting This is 10% and the maximum specific gravity can be defined as 70%. The service policy may include information defining a regional weight limit of assets included in the portfolio. For example, the weight limit of domestic equity-type assets included in the portfolio, the weight limit of overseas equity-type assets, the weight limit of domestic bond-type assets, and the weight limit of overseas bond-type assets may be defined as service policies.

In addition, the service policy may include information defining the number of assets constituting the portfolio and an investment target type of each asset.

Table 5 below shows an example in which the investment target type of each asset constituting the portfolio is defined.

fund # Investment target type fund 1 domestic equity fund fund 2 Overseas Advanced Equity Fund fund 3 Overseas Emerging Equity Fund fund 4 bond fund

Taking Table 5 as an example, the portfolios for asset management services include one domestic stock-type fund, one overseas advanced stock-type fund, one overseas emerging stock-type fund, and one bond-type fund. can be The service policy defined as described above may be applied as a constraint when configuring a portfolio for each strategy in the asset management server 200 to be described later or configuring a final portfolio through strategy synthesis. For example, the asset management server 200 limits the risk grade of each asset constituting the portfolio to satisfy the risk limit according to the investment propensity of the customer when composing the portfolio for each strategy or composing the final portfolio through strategy synthesis. can do. In addition, for example, the asset management server 200 configures the portfolio to satisfy the weight limit for each asset type of the portfolio according to the customer's investment propensity when configuring the portfolio for each strategy or composing the final portfolio through strategy synthesis. You can limit the asset type of each asset.

Again, referring to FIG. 1 , the customer information DB 130 includes the customer's personal information (eg, the customer's name, resident registration number, address, phone number, mobile phone number, e-mail address, etc.), the customer's member information (eg, , member ID information or CIF number provided in the customer ledger provided on the asset management service system in response to the customer signing up as a financial transaction customer of a financial institution), customer asset management information, customer investment propensity Information can be stored and managed.

The asset management server 200 is a server for providing an asset management service to a customer, and may provide the customer with an asset allocation model portfolio suitable for the customer's investment propensity. The asset management server 200 may include a screening unit 210 , a portfolio creation unit 220 for each strategy, and a strategy synthesis unit 230 .

The screening unit 210 reads the fund list (hereinafter, referred to as 'all fund pool') and asset information of all funds sold by the corresponding financial institution from the financial product DB 110, and based on this, reads from the financial institution Screening can be performed on the entire fund pool being sold. In addition, the screening unit 210 selects candidate funds suitable for constructing a portfolio of an asset allocation model from the entire fund pool through screening, and a fund list (hereinafter, 'portfolio fund') in which the selected candidate funds are abbreviated for each major asset group. pool') can be created. Here, the major asset class is a classification of funds sold by financial institutions by investment type, and may include domestic stock types, overseas advanced stock types, emerging stock types, and bond types (overseas bond types, domestic bond types), and the like.

The screening unit 210 may use various screening methods when selecting a candidate fund suitable for each major asset class.

For example, before selecting a candidate fund, the screening unit 210 may perform an operation of excluding an overheated fund from selection of a candidate fund by using a relative strength index (RSI). Here, the RSI calculates the average value of the amount of change and the amount of rice that has fallen compared to the previous day's price for a certain period of time, and if the increase is large, it is overbought, and if the decrease is large, it is oversold. Equation 1 and are calculated together.

[Equation 1]

In Equation 1 above, RS is calculated by Equation 2 below.

[Equation 2]

The screening unit 210 determines the overbought phase if the RSI is 70 to 75% or more, and determines the oversold phase if the RSI is 25 to 30% or less. , it may be judged as an overheated fund and excluded from the candidate fund.

After that, the screening unit 210 determines that, among the remaining funds, a fund with a low one-month rate of return below a predetermined value (eg, the bottom 50%) is a fund with weak momentum in the short term, and excludes it from the candidate fund. , from among the remaining funds, at least one fund selected in the order of the highest 3-month rate of return for each major asset class may be selected as a candidate fund. That is, among the remaining funds, excluding the overheated fund and the plunging fund, a fund whose trend is alive in the medium term can be selected as a candidate fund for each major asset class.

When selecting a candidate fund for each asset group through screening, the screening unit 210 may select a candidate fund for each risk grade. For example, if the risk grades of funds belonging to each asset class are classified from grade 1 to grade 3, the risk grade for each asset group (e.g., domestic stock type, overseas advanced stock type, emerging stock type, bond type) through screening These three different candidate funds can be selected.

The portfolio creation unit 220 for each strategy configures a plurality of fund combinations using the funds included in the portfolio fund pool, and applies a plurality of different strategies to the plurality of fund combinations configured in this way to determine the asset proportions of each fund combination. Portfolios can be created by making decisions. Here, each portfolio may include a corresponding fund combination and an asset weight of each fund included in the fund combination. The asset class (investment target type) of the funds constituting the portfolio and the number of funds may be defined by the above-described service policy. Taking Table 5 as an example, each portfolio may be composed of a total of four funds: one domestic stock-type fund, one overseas advanced stock-type fund, one emerging stock-type fund, and one bond-type fund.

The plurality of strategies used for portfolio creation in the portfolio creation unit 220 for each strategy are strategies for determining the asset weight of each fund included in the portfolio, and have different investment objectives, and thus different parameters for portfolio construction can be used In the following description, a case of using Momentum (MoM), Black-Litterman (BL), and Risk Parity (RP) as a plurality of strategies will be described as an example. However, the strategy used to configure the portfolio of the asset allocation model in the asset management server 200 is not limited thereto, and various investment strategies may be used in addition to the above-described strategies.

Table 6 below outlines the characteristics of each strategy.

strategy investment purpose characteristic note momentum Maximize your returns Tracking the momentum of recent asset prices Reflecting AI market forecast values Black Ritterman The balance between maximum return and minimum volatility Pursuing high return on risk Market capitalization by major asset, reflecting the view of investment experts risk parity Minimize volatility Diversify risk between assets as much as possible Does not reflect predicted values, uses historical distribution

Below, each strategy will be described in detail with reference to Table 6 above. The momentum strategy is a strategy that aims to maximize the expected return of the entire portfolio by maximizing the ratio of the returns of each fund constituting the portfolio of the asset allocation model. This momentum strategy may show a risky move in which the range of decline in a down market is larger than that of other strategies, as the range of ups and downs is larger than that of other strategies. The momentum strategy can be theoretically defined through Equation 3 below.

[Equation 3]

In Equation 3 above, E[R _P ] represents the expected return of the entire portfolio, and E[R _{1 … n} ] represents the expected return for each asset, W _{1… nt} represents the weight of each individual asset.

In Equation 3 above, the _{weight of each asset with the maximum expected return E[RP} ] may be the final result of the weight of each asset in the portfolio finally constructed through the momentum strategy.

Next, the Black Literman strategy aims to maximize the return relative to the expected volatility of the portfolio by maximizing the ratio of the ratio of the returns of each fund constituting the portfolio of the asset allocation model by the volatility of the portfolio. am. This Black Literman strategy is characterized by stable movement.

In the Black Literman strategy, the weight of each investment asset (fund) is calculated by market capitalization, and the risk coefficient (λ) is calculated based on this. (Implied Equilibrium Return(Π)) is calculated.

[Equation 4]

At this time, since the market capitalization of foreign stocks/bonds is much higher than that of domestic stocks/bonds, the weight of each market capitalization is used as it is, and the value input from the investment expert through the manager terminal 300 is used. have.

In Equation 4 above, E[R _m ] is the expected return of the market, R _f is the risk-free interest rate, W _mkt is the proportion of market capitalization proportional to the investment asset group, and ∑ represents the covariance of the return on the investment asset group. .

In addition, in the Black Literman strategy, qualitative predictions for each asset can be quantified as shown in Equation 5 below, and can be reflected in the rate of return for each asset.

[Equation 5]

In Equation 5 above, Ω represents the rate of return for each asset that reflects the predicted value, P is the determinant that reflects the predicted value, and ∑ represents the covariance of the return of the investment asset group.

In the Black Literman strategy, by combining the opinions determined through Equations 4 and 5 above, the weight (W _BL ) for each asset may be finally determined as shown in Equation 6 below.

[Equation 6]

The risk contribution (MRC: Marginal Risk Contribution) used in this risk parity strategy refers to the risk (volatility) of the portfolio that increases when the weight of a specific stock is increased by one unit, all other conditions being the same. It can be calculated as in Equation 7 below.

[Equation 7]

In Equation 7 above, MRC _i represents the level of risk (volatility) that asset i contributes to the overall portfolio risk (volatility), W _i represents the proportion of asset i in the entire portfolio, and σ[R _p ] is the portfolio represents the level of risk (volatility) of

When the risk contribution for each asset calculated through Equation 7 above is calculated, as shown in Equation 8 below, the sum of the risk contribution for each asset constituting the portfolio is calculated, that is, the risk contribution of the portfolio.

[Equation 8]

In Equation 8 above, the proportion of each asset with the lowest risk contribution may be the final result of the proportion of each asset in the portfolio finally constructed through the risk parity strategy.

The portfolio creation unit 220 for each strategy configures a plurality of portfolios corresponding to each strategy from the portfolio fund pool through the MoM module 221 , the BL module 222 , and the RP module 223 , and among them, each strategy You can choose the best portfolio for you.

That is, the MoM module 221 generates at least one portfolio from the portfolio fund pool by using the momentum strategy, and selects the portfolio with the greatest expected return for a predetermined period (eg, 3 months) from among the generated portfolios in the momentum strategy. of the optimal portfolio (hereinafter referred to as a 'momentum portfolio').

In addition, the BL module 222 generates at least one portfolio from the portfolio fund pool using the Black Literman strategy, The portfolio in which the ratio of the ratio of returns divided by the volatility of the portfolio is the maximum can be selected as the optimal portfolio in the Black Literman strategy (hereinafter referred to as the 'Black Literman Portfolio').

In addition, the RP module 223 generates at least one portfolio from the portfolio fund pool using the risk parity strategy, and risks the portfolio with the minimum volatility for a predetermined period (eg, 3 months) among the generated portfolios. An optimal portfolio in the parity strategy (hereinafter, referred to as a 'risk parity portfolio') may be selected.

Meanwhile, the portfolio creation unit 220 for each strategy may refer to the investment propensity information of the customer and the above-described service policy in order to configure a plurality of portfolios from the portfolio fund pool by using the plurality of strategies. Customer's investment propensity information and service policy may act as a constraint in the process of constructing a portfolio using multiple strategies.

That is, when the portfolio creation unit 220 for each strategy configures a plurality of portfolios from the portfolio fund pool using a plurality of strategies, the investment propensity (aggressive investment type, active investment type, risk neutral type, stability seeking type) according to the service policy etc.) defined risk limits (minimum risk grade and maximum risk grade), weight limit by asset type, etc., may limit the risk grade of funds included in the portfolio and weight by asset type. For example, in the case of a portfolio provided to an aggressive investment type customer, the type of fund included in the portfolio and the proportion of assets by each fund may be limited so that the overall risk rating of the portfolio is limited to a minimum of 2 and a maximum of 4. Also, for example, in the case of a portfolio offered to aggressive investment clients, among the assets constituting the portfolio, stocks comprise at least 50%, bonds at least 25%, and risky assets at least 25% and maximum. 65%, up to 30% of assets with risk grade 2 or higher, up to 60% of domestic stocks, up to 50% of foreign stocks, up to 25% of emerging stocks, and minimum of individual funds It is possible to limit the weight of each fund included in the portfolio so that the weighting is 10% and the maximum weighting is limited to 70%.

In addition, when constructing a plurality of portfolios from a portfolio fund pool using a plurality of strategies, the portfolio creation unit 220 for each strategy calculates the proportion of domestic assets and foreign assets based on the regional weight limit of assets defined by the service policy. may be limited.

The strategy synthesis unit 230 configures the optimal asset allocation model portfolio (momentum portfolio, Black Literman portfolio, and risk parity portfolio) for each strategy by the strategy-specific portfolio creation unit 220, By synthesizing them based on , a portfolio of the final asset allocation model can be created.

Model Asset Allocation Model purpose input objective definition multi strategy - Reference period by fall: 90 days (3 months)
- Daily rate of return Max(α × return - volatility) - Genetic algorithm related variables: mutation rate (5%), number of mutations (20), number of generations (3 generations) MAX (correlation coefficient)

In order to synthesize the momentum portfolio, the Black Literman portfolio, and the risk parity portfolio, the strategy synthesizing unit 230 may first perform a process of calculating an optimal weight between strategies. That is, the strategy synthesis unit 230 considers all returns and volatility of the momentum portfolio, the Black Literman portfolio, and the risk parity portfolio for a predetermined reference period (eg, 90 days or 3 months), Max(α × It is possible to obtain an optimal weight that satisfies the rate of return-volatility). Here, the rate of return represents the average rate of return over the reference period, and the volatility represents the volatility over the reference period. In this way, if the scale between the return and the volatility (eg, 1% return, 5% volatility) and the fluctuation range are significantly different, the effect can be offset by multiplying the return by a certain factor. In addition, the weight of the average return relative to volatility is adjusted through risk preference (α, for example, α=10), and additional weight can be placed on the return to maximize return versus volatility. As a result, bull market (usually a phase in which returns are rising, choosing a strategy with high returns to make money in a bull market) and a bear market (usually phase in which volatility increases, choosing a strategy with low volatility to defend in a bear market) The asset allocation model can be designed so that both sides can respond.

Meanwhile, the funds selected for portfolio composition may be different for each strategy. Table 8 below shows an example of an optimal portfolio for each strategy generated by the portfolio creation unit 220 for each strategy.

strategy Asset Share Fund name bond domestic
stock Overseas
advanced emerging bond domestic
stock Overseas
advanced emerging MoM 27% 24% 25% 24% Fund A Fund B Fund C Fund E BL 34% 24% 26% 16% Fund A fund F Fund C Fund G RP 45% 10% 35% 10% Fund H fund F Fund C Fund E

Taking Table 8 above as an example, in the case of momentum (MoM) strategy, bond-type fund A, domestic stock-type fund B, overseas advanced fund C, and emerging fund E were selected to form the portfolio, and their respective proportions were 27% and 24 %, 25%, and 24%. In addition, in the case of the Black Literman (BL) strategy, bond-type fund A, domestic stock-type fund F, overseas advanced fund C, and emerging fund G were selected to form the portfolio, and their respective weights were 34%, 24%, and 26 %, and 16%. In addition, in the case of risk parity (RP) strategy, bond-type fund H, domestic stock-type fund F, overseas advanced fund C, and emerging fund E were selected to form the portfolio, and their respective weights were 45%, 10%, and 35%. , and 10%. Therefore, there are a total of 7 funds (Funds A to G) selected by the momentum, Black Literman, and risk parity strategies, 2 bond-type funds, 2 domestic equity-type funds, 1 overseas advanced equity-type fund, and 3 emerging funds contains a dog. The portfolio of the asset allocation model finally provided to the customer may include one recommended fund for each of a plurality of major asset classes, as defined by the above-described service policy. However, as described above, the funds constituting the momentum portfolio, the Black Literman portfolio, and the risk parity portfolio may be different from each other. It is necessary to carry out the process of deriving That is, among the multiple funds included in the momentum portfolio, the multiple funds included in the Black Literman portfolio, and the multiple funds included in the risk parity portfolio, in the final portfolio, each major asset class (domestic stock type, overseas advanced stock type, It is necessary to select a recommended fund to represent the emerging stock type and bond type).

The strategy synthesizer 230 may use a genetic algorithm (GA) for this purpose. Hereinafter, with reference to FIG. 2 , a method for selecting a recommended fund to represent each major asset class in a recommended portfolio through a genetic algorithm will be described. 2 illustrates an example of selecting a fund combination included in a final portfolio through a genetic algorithm.

Referring to FIG. 2 , the strategy synthesis unit 230 configures combinations of all funds (fund combination 1, fund combination 2, ?? fund combination m) selected for each strategy in order to apply the genetic algorithm. Taking Table 8 as an example, the combination of 2 bond-type funds, 2 domestic equity-type funds, 1 overseas advanced-type fund, and 3 emerging funds, selected by momentum, Black Literman, and risk parity strategies, is 2 (bond type) × 2 (domestic stock type) × 1 (foreign advanced stock type) × 3 (emerging type), that is, a total of 12 fund combinations can occur.

The strategy synthesis unit 230 applies a genetic algorithm to the fund combinations thus generated. At this time, the asset proportion obtained by the (return-volatility) method is applied to the initial value of the asset proportion of each fund combination applied in the 0th generation of the genetic algorithm.

The strategy synthesis unit 230 generates mutations by changing the proportion of assets of each fund combination to a certain level from the proportion of assets of the previous generation whenever a generation is changed by the genetic algorithm (eg, generation 0 -> generation 1). do. Taking Table 7 as an example, 20 variations can be created by changing the weight of assets at a variation rate of approximately 5% each time a generation changes.

When variations of asset weight combinations are generated, the strategy synthesis unit 230 measures the correlation coefficient of each variation and whether the constraint is satisfied, and the asset weight combination that does not satisfy the constraint condition is eliminated. In addition, if the asset weighting of the fund combination satisfies the constraint, at least one weighting combination selected in the order of the highest correlation coefficient is selected, and the remaining asset weighting combinations are culled. Here, the constraint means a constraint defined by the above-described service policy. In addition, the correlation coefficient is based on the optimal weight for each strategy calculated in the above-described manner in the strategy synthesis unit 230, and the theoretical standard value by the asset allocation model that theoretically combines the momentum portfolio, the Black Literman portfolio, and the risk parity portfolio; It shows the similarity between the base prices of the asset allocation model that combines the corresponding funds based on the combination of the weight loss ratio generated through the variation of the genetic algorithm. Here, the asset allocation model, which theoretically combines the momentum portfolio, the Black Literman portfolio, and the risk parity portfolio, includes all the funds constituting the momentum portfolio, the Black Literman portfolio, and the risk parity portfolio, each of which is allocated to itself. It can be obtained by combining the asset proportion and the optimal proportion for each strategy calculated by the strategy synthesis unit 230 according to the given asset proportion.

The strategy synthesis unit 230 repeatedly performs this process for each fund combination until n generations are reached, and selects the asset proportion combination with the highest correlation coefficient among the asset proportion combinations generated by repeating until the n generation of the corresponding fund. is selected as the final specific gravity combination of And, the fund combination with the highest correlation coefficient of the final selected asset weight combination among the plurality of fund combinations (fund combination 1, fund combination 2, ?? fund combination m), and its asset weight combination, the fund combination and asset of the final portfolio Select by weight combination.

3 is an asset allocation model theoretically synthesized by a plurality of portfolios for each strategy and a performance trend and correlation coefficient for each investment propensity of an asset allocation model that is finally provided to customers by the asset management server 200 according to the embodiment. , 3, the final portfolio derived from the portfolios for each strategy through the genetic algorithm has a correlation coefficient of 90% or more with the asset allocation model theoretically synthesizing the portfolios for each strategy according to the optimal weight. high can be seen. That is, in the embodiment, an asset allocation model portfolio having performance similar to that of an asset allocation model theoretically synthesized by a plurality of strategy-specific portfolios according to an optimal weight may be generated through a genetic algorithm.

As described above, when the final portfolio is determined through the genetic algorithm, the asset management server 200 delivers it to the window terminal 400 or the client terminal 500 to provide it to the customer.

Hereinafter, a method of providing an asset management service according to an embodiment will be described with reference to FIG. 4 .

4 schematically illustrates a method of providing an asset management service according to an embodiment, and illustrates a method of providing an asset management service for providing a portfolio of an asset allocation model to a customer. The method of providing the asset management service of FIG. 4 may be performed by the asset management server 200 described with reference to FIG. 1 .

Referring to FIG. 4 , in order to configure the customer's asset allocation model portfolio, the asset management server 200 first creates a portfolio fund pool through screening of the entire fund pool ( S11 ). That is, the asset management server 200 selects candidate funds suitable for composing the portfolio of the asset allocation model from the entire fund pool through screening, and creates a portfolio fund pool in which the selected candidate funds are abbreviated for each of a plurality of major asset groups.

In the step S11, the asset management server 200 before selecting a candidate fund, the fund determined to be an overheated fund based on the relative strength index (RSI), and the one-month rate of return is lower than a predetermined value (for example, The bottom 50%) fund is excluded from the candidate fund, and from the remaining funds, screening is performed by selecting at least one fund selected in the order of the highest 3-month return for each major asset class as a candidate fund. can

When the portfolio fund pool is created, the asset management server 200 creates a portfolio for each strategy from the portfolio fund pool through a plurality of different strategies (S12).

In step S12, the asset management server 200 configures a plurality of fund combinations using the funds included in the portfolio fund pool, and applies a plurality of different strategies to each of the plurality of fund combinations configured in this way. Portfolios are created by determining asset weights, and the optimal portfolio for each strategy can be selected from among them. For example, the asset management server 200 generates at least one portfolio from the portfolio fund pool using the momentum strategy, and selects the portfolio with the highest expected return for a predetermined period (eg, 3 months) among the generated portfolios. The final choice is the portfolio of momentum strategies (momentum portfolio). Also, for example, the asset management server 200 generates at least one portfolio from the portfolio fund pool using the Black Literman strategy, and each of the generated portfolios for a predetermined period (eg, 3 months) The portfolio in which the weight sum of the returns of assets (funds) is the maximum divided by the volatility of the portfolio can be finally selected as the portfolio of the Black Literman strategy (Black Literman portfolio). Also, for example, the asset management server 200 generates at least one portfolio from the portfolio fund pool using a risk parity strategy, and the volatility for a predetermined period (eg, 3 months) among the generated portfolios is The smallest portfolio can be finally selected as the portfolio of risk parity strategy (risk parity portfolio).

As described above, when a portfolio for each strategy is configured, the asset management server 200 calculates an optimal weight between a plurality of strategies based on the yield and volatility of the portfolio for each strategy ( S13 ). Then, an optimal fund combination and asset ratio are determined from the funds constituting the portfolios for each strategy through a genetic algorithm (S14).

In the step S14, the asset management server 200 configures a plurality of fund combinations from a plurality of funds constituting a plurality of strategy-specific portfolios. And, through a genetic algorithm, among a plurality of fund combinations configured in this way, a fund combination with a high correlation coefficient with the result of combining a plurality of portfolios for each strategy according to the optimal weight calculated in step S13, and the asset ratio thereof to the optimal fund combination and It is determined by the proportion of assets.

When the optimal fund combination and asset weight are determined through the step S14, the asset management server 200 generates a final portfolio of the asset allocation model based on this (S15), and provides it to the customer (S16).

5A and 5B are diagrams for explaining the effect of a method for providing an asset management service according to an embodiment, and FIG. 5A is a diagram illustrating a change in the proportion of each strategy calculated according to the method for providing an asset management service according to the embodiment. , FIG. 5B is an asset allocation model portfolio (see final MP in FIG. 5B ) finally provided to customers by the asset management service according to the embodiment, and asset allocation configured by each single strategy (momentum, blacklitermann, risk parity). The performance trend is shown by the theoretical synthesis of the model portfolios and the portfolios for each strategy based on the optimal weight (refer to the 'theoretical standard price' in FIG. 5B).

5A and 5B , an asset allocation model theoretically synthesizing portfolios constructed by a plurality of strategies based on an optimal weight and a portfolio finally provided to customers by the asset management service according to the embodiment have similar performance It can be seen that both of them performed better than portfolios constructed using a single strategy.

The above-described embodiment can be implemented as computer-readable code on a medium in which a program is recorded. The computer-readable medium includes all types of recording devices in which data readable by a computer system is stored. Examples of computer-readable media include Hard Disk Drive (HDD), Solid State Disk (SSD), Silicon Disk Drive (SDD), ROM, RAM, CD-ROM, magnetic tape, floppy disk, optical data storage device, etc. There is also a carrier wave (eg, transmission over the Internet) that is implemented in the form of. In addition, the computer may include a control unit of the terminal. Accordingly, the above detailed description should not be construed as restrictive in all respects but as exemplary. The scope of the present invention should be determined by a reasonable interpretation of the appended claims, and all modifications within the equivalent scope of the present invention are included in the scope of the present invention.

Claims (18)

A screening unit that selects candidate funds for each major asset class through screening of funds sold by financial institutions;
a portfolio creation unit for each strategy that generates a plurality of portfolios respectively corresponding to the plurality of strategies from the candidate funds by using a plurality of different strategies, and
A plurality of fund combinations are formed from a plurality of funds constituting the plurality of portfolios, and the plurality of portfolios are theoretically synthesized based on the optimal ratio between the plurality of strategies determined according to the yield and volatility of the plurality of portfolios. Acquire an asset allocation model, and based on a genetic algorithm, determine a fund combination and asset ratio with a high correlation coefficient with an asset allocation model theoretically synthesizing the plurality of portfolios among the plurality of fund combinations as the final fund combination and final asset ratio and a strategy synthesis unit for generating a final portfolio of an asset allocation model according to strategy synthesis according to the final fund combination and the final asset weight,
The plurality of strategies are strategies for determining the asset weight of each fund included in the corresponding portfolio, and include a momentum strategy, a Black-Litterman strategy, and a risk parity strategy. , an asset management server. According to claim 1,
The strategy synthesis unit,
An asset management server for synthesizing the plurality of portfolios by combining all the funds constituting the plurality of portfolios according to an asset weight given by a combination of the optimal weight and the weight of assets allocated to each fund in each of the plurality of portfolios. 3. The method of claim 2,
The strategy synthesizing unit determines the asset weights of each of the plurality of fund combinations through the genetic algorithm, and the asset weighting with the highest correlation coefficient with the result of synthesizing the plurality of portfolios among the asset weights of each of the plurality of fund combinations The asset management server that determines the final asset weight. According to claim 1,
The correlation coefficient is an asset management server corresponding to the degree of similarity between the reference values. According to claim 1,
Each of the plurality of portfolios and the final portfolio includes recommended funds selected for each of a plurality of major asset classes, and an asset weighting of the recommended funds. 6. The method of claim 5,
The screening unit may include at least one fund determined to be an overheated fund based on a relative strength index among the funds sold by the financial institution and a fund with a low return rate of less than or equal to a predetermined value for one month from among the remaining funds. , an asset management server for selecting at least one fund selected in an order of high yield for each of the plurality of major asset groups as the candidate fund. delete According to claim 1,
It further includes a service policy database for storing a service policy in which a risk limit and a weight limit for each asset type are defined for each investment type,
The strategy-specific portfolio creation unit is an asset management server for limiting risk grades and asset types of funds constituting the plurality of portfolios based on the service policy. According to claim 1,
The optimal weight is an asset management server calculated by a difference between a value obtained by multiplying an average return for a predetermined period by a risk preference and volatility during the predetermined period. A method of providing an asset management service of an asset management server, the method comprising:
Selecting candidate funds for a plurality of major asset classes through screening of funds sold by financial institutions;
determining, from the candidate funds, a plurality of portfolios each corresponding to the plurality of strategies, using a plurality of different strategies;
determining an optimal weight between the plurality of strategies based on returns and volatility of the plurality of portfolios;
Based on a genetic algorithm, among a plurality of fund combinations combined from a plurality of funds constituting the plurality of portfolios, an asset allocation model theoretically synthesizing the plurality of portfolios according to the optimal weight and a fund combination with a high correlation coefficient and determining the asset weighting as the final fund combination and final asset weighting;
generating a final portfolio of an asset allocation model according to strategy synthesis according to the final fund combination and the final asset weight; and
providing said final portfolio to a customer;
The plurality of strategies are strategies for determining the asset weight of each fund included in the corresponding portfolio, and include a momentum strategy, a Black-Litterman strategy, and a risk parity strategy. , how to provide asset management services. 11. The method of claim 10,
The method further comprising the step of synthesizing the plurality of portfolios by combining all the funds constituting the plurality of portfolios according to the asset weight given by the combination of the optimal weight and the weight of assets allocated to each fund in each of the plurality of portfolios How to provide asset management services. 12. The method of claim 11,
The step of determining the final fund combination and the final asset proportion,
determining asset weights of each of the plurality of fund combinations through the genetic algorithm; and
and determining an asset proportion having the highest correlation coefficient with a result of synthesizing the plurality of portfolios among the asset proportions of each of the plurality of fund combinations as the final asset proportion. 11. The method of claim 10,
The correlation coefficient is a method of providing an asset management service corresponding to a degree of similarity between reference prices. 11. The method of claim 10,
Each of the plurality of portfolios and the final portfolio includes recommended funds selected for each of the plurality of major asset groups, and asset ratios of the recommended funds. 11. The method of claim 10,
Selecting the candidate funds comprises:
Among the funds sold by the financial institution, excluding at least one fund determined to be an overheated fund based on a relative strength index and a fund having a low yield for one month or less, among the remaining funds, the plurality of A method of providing an asset management service comprising the step of selecting at least one fund selected in the order of highest return for each major asset class as the candidate fund. delete 11. The method of claim 10,
The step of determining the plurality of portfolios includes limiting the risk class and asset type of the funds constituting the plurality of portfolios based on a service policy in which a risk limit and a weight limit for each asset type are defined for each investment type How to provide asset management services. 11. The method of claim 10,
The step of calculating the optimal specific gravity is,
and calculating the optimal weight based on a difference between a value obtained by multiplying an average return for a predetermined period by a risk preference and volatility for the predetermined period.

Images