JP4828953B2 - System and method for supporting analysis of asset management outcomes - Google Patents

Description

  The present invention relates to a technique for supporting analysis of asset management results.

  For example, it is known to calculate a return by the modified Dietz method in order to obtain a result of asset management.

http://www.gpif.go.jp/unyou/yougoshuu.html

  By the way, as a method for supporting the analysis of the result of asset management, there is a method of dividing the result for each factor. Dividing the results of asset management into factors is called “factor decomposition”. In general, it is possible to separately perform (A) and (B) of (A) factor decomposition into sector tilt effect and issue selection effect, and (B) factor decomposition into initial holdings and trading during the period. is there.

  The analysis of asset management results should be as accurate as possible. For that purpose, it is considered that the result of asset management can be decomposed into more detailed factors than the above method (A) or (B).

  In addition, if an apparatus capable of performing factorization is realized, it is desirable that the apparatus is not only capable of performing factorization but also more convenient for the user.

  Therefore, an object of the present invention is to provide an asset management result analysis support technology that can decompose the results of asset management into more detailed factors than in the past and is more convenient than simply performing factor decomposition.

  Other objects of the present invention will become clear from the following description.

  The asset management result analysis support system according to the present invention comprises a reading means, an overall rate of return calculation means, a discrimination means, and a factor decomposition means. The reading unit reads a plurality of pieces of change data related to inflow or outflow of assets from the storage device. The overall rate of return calculation means calculates the overall rate of return using a plurality of read transfer data. The discriminating unit discriminates whether each read change data is used for calculating the initial holding effect or the interim purchase effect. Factor decomposition means breaks down the results of asset management into three types of effects: initial holding effect, purchase effect during the period, and sale effect during the period. The output means outputs information based on the result of the factorization.

  When the factor decomposition means has interest / dividend date of a stock that is held or purchased belongs to the calculation target period, based on the interest / dividend data of the stock and the result of discrimination, the initial holding effect and the The purchase effect is calculated, and then the interim sale effect is calculated by subtracting the initial holding effect and the interim purchase effect from the calculated overall rate of return.

  As a result, even if the interest / dividend date of the sold brand belongs to the calculation target period, the interest / dividend data of the brand can be made unnecessary in order to calculate the effect of sale during the period.

  In the first embodiment, the overall rate of return calculation means can update the total balance of assets in the calculation target period in a predetermined storage area based on each read change data. The factor decomposition means can perform the factor decomposition using the total balance of assets prepared by the overall rate of return calculation means.

  In the second embodiment, each time the change data is read out, the means for writing the read change data in a predetermined storage area and the factor decomposition means using the transfer data written in the predetermined storage area And a means for deleting the transfer data used for the update from the storage area each time the update is completed.

  The asset management result analysis support method according to the present invention includes a step of reading a plurality of change data relating to inflow or outflow of assets from a storage device, and a step of calculating an overall rate of return using the plurality of read change data. , A step to determine whether each read-out change data is used for initial holding calculation or interim purchase calculation, and asset management results are classified into three types: initial holding effect, interim purchase effect, and interim sale effect Factorizing the effect, and outputting information based on the factorization result. In the factor decomposing step, when the interest / dividend date of the stock that is held or purchased belongs to the calculation target period, the initial holding effect and the above-mentioned effect based on the interest / dividend data of the stock and the result of the determination The interim purchase effect is calculated, and then the interim sale effect is calculated by subtracting the initial holding effect and the interim purchase effect from the calculated overall rate of return.

  Specifically, in this method, for example, in order to calculate the overall rate of return using the first step of reading the transfer data from the storage device during the calculation target period and the plurality of read transfer data With reference to the second step of updating the first value of the required type on the predetermined storage area and the predetermined data element of each read out read data, each transfer data is calculated for the initial holding effect and The third step for determining which of the purchases is used for the calculation of the interim purchase, and the second value of the type corresponding to the result of the determination among the multiple types of values necessary for factorization of the asset management results The fourth step of updating on the predetermined storage area can be repeated until all the different data related to the calculation target period is read out. A step of calculating an overall rate of return using a first value after the second step after all the dissimilar data are read, and the first after all the dissimilar data are read. And factorization using the second value after the four steps. In the third step, when the interest / dividend date of the stock that is held or purchased belongs to the calculation target period, the second value is calculated using the interest / dividend data of the stock. It can be updated on the storage area. In the factor decomposition step, the initial holding effect and the interim purchase effect can be calculated using the updated second value. The “change data related to the calculation target period” includes at least change data within the calculation target period. The “change data relating to the calculation target period” may include the last change data of the immediately preceding period for each brand in order to calculate the initial holding effect.

  By the way, the storage device can be, for example, a disk-type storage medium (for example, a hard disk or a DVD (Digital Versatile Disk)), or a disk drive (for example, a hard disk drive) provided with a disk-type storage medium. The predetermined storage area may be an area provided on a storage device different from the storage device, for example, a memory.

  The “change” can include, for example, a change in funds due to interest or dividends in addition to a change in buying or selling stock.

  The “interest / dividend data” may be mixed with other types of transfer data as one type of transfer data in one database, or may exist in a database different from the database in which the transfer data exists. May be.

  The asset management result analysis support system may be a single computer machine or a computer system constructed by a plurality of computer machines. The asset management result analysis support system can be applied to, for example, a server of a server client system, and is not limited to this, and can be applied to other systems.

  Each means of the asset management result analysis support system can be realized by hardware, a computer program, or a combination thereof. For example, the asset management result analysis support system may be constructed by installing the computer program from a storage medium such as a CD-ROM or by downloading the computer program via a communication network. The asset management result analysis support system includes a processor (for example, CPU), and the processor can read and execute the computer program.

  According to the present invention, it is possible to provide an asset management result analysis support technology that is capable of decomposing asset management results into more detailed factors than before and that is more convenient than simply performing factor decomposition.

  Hereinafter, a server machine (hereinafter referred to as a server) to which an asset management result analysis support system according to an embodiment of the present invention is applied will be described.

  The server according to the present embodiment can decompose the results of asset management into more detailed factors than before, in other words, can perform improved factor decomposition.

  In addition, the improvement factor decomposition can calculate three types of effects: initial holding effect, purchase effect during the period, and sale effect during the period. In order to calculate various effects, stocks whose interest dividend date is within the calculation target period can be calculated. Interest / dividend data in which interest / dividend date and interest / dividend date are recorded is required. However, according to the present embodiment, with regard to the effect of sale during the period, the interest / dividend date belongs within the calculation target period. Even so, the interest / dividend data can be made unnecessary (this advantage will be described in detail later).

  In short, according to the present embodiment, it is possible to provide asset management result analysis support technology that can decompose the results of asset management into more detailed factors than before and that is more convenient than simply performing factor decomposition. It is.

  First, in order to make it easy to understand the description of the server according to the present embodiment, conventional factor decomposition and improvement factor decomposition will be described. At this time, in order to make the explanation easy to understand, it is assumed that the asset to be purchased or sold is a stock.

1. About factor decomposition in the past In general, in the asset management industry such as stock investment, a person who deposits funds from an investor (called an investment manager) uses the funds to make an advantageous classification (hereinafter referred to as a sector). ) Buy or sell stocks of advantageous brands, and manage assets so that profits are greater than the initial investment. In this case, the profits that can be obtained depend greatly on how the managers manage their assets.

  This method of asset management is based on the choice of diversified investment in which stock in which sector and how much weight to invest (the asset of this advantageous diversified investment is called a portfolio), and at what time It has a speculative aspect of whether it was sold or held without being purchased or sold. Asset management can be evaluated from these two aspects.

  When profits or losses from asset management (collectively, asset management results) are generated, analyzing the factors that generated the asset management results can be done by examining future asset management policies or managing the performance of managers. It has an important meaning to evaluate (good or bad).

  The results of asset management (hereinafter also referred to as excess returns) are generally calculated based on the asset valuation at the beginning of the period, the asset valuation at the end of the period, and cash flows during the period. Here, the beginning of the term is the beginning of the calculation period, the end of the term is the end of the calculation period, and the assets are actually stocks and bonds. The calculation period refers to the period for measuring the performance of an investment divided into periods of a predetermined length for continuous investments.

In order to obtain the result of this asset management, the return r is calculated by the following modified Diets method for each asset, that is, the following equation (10),
_{r = (V 1 -V 0 -Σ} t C t) / (V 0 + Σ t C t × (1-t)) ... (10)
Is being used.

Here, it is assumed that the length of the period for measuring the investment performance is 1. V ₀ and V ₁ are the market value of the portfolio at the beginning and end of the period, respectively, and C _t is the cash flow generated at time t (0 <t <1). This C _t is positive for the inflow of funds into the portfolio (purchase of assets) and negative for outflow of funds from the portfolio (sale of assets).

  By comparing the return r based on the above-mentioned modified Diets method for each asset with the return R of the benchmark index that is the basis for revenue, judging whether the asset management results are good or bad by the magnitude of the excess return (r-R) Can do. The return R of this benchmark index conceptually refers to profits obtained without special asset management, such as interest on deposits, etc. In the world of Japanese stock investment, the TSE stock price including dividends The index (TOPIX) is often used as the benchmark index return R.

  There are the following two methods for factoring the excess return (r-R).

  The first method is a method of factorizing asset management results into sector tilt effect and stock selection effect. Here, the sector means one set of investment objects, and in the world of stock investment, it is normal to refer to an industry as described above. The sector tilt effect refers to the effect of changing the sector type and weight of each sector in the portfolio. The brand selection effect refers to an effect obtained by eliminating brands that cannot obtain profits within the same sector and by selecting brands that can obtain a large profit.

According to the first method, the above formula (10) is replaced by the following formula (11):
r = {Σ _s (V _s1 −V _s0 −Σ _t C _st )} / {V ₀ + Σ _t C _t × (1−t)}
= Sigma _{s [{(V s0 + Σ t C} st × (1-t)) / (V 0 + Σ t C t × (1-t))}
_{× {(V s1 -V s0 -Σ} t C st) / (V s0 + Σ t C st × (1-t))} ] ... (11)
It transforms as follows.

Here, V _s1 , V _s0 , and C _st are V ₁ , V ₀ , and C _t in each sector s. That is, V ₁ = Σ _s V _s1 , V ₀ = Σ _s V _s0 , and C _t = Σ _s C _st are established.

Then, in the above _{(11), {(V s0 + Σ t} C st × (1-t)) / (V 0 + Σ t C t × (1-t))} , and investment weights x to sector s and _s, as _{{(V s1 -V s0 -Σ t} C st) / (V s0 + Σ t C st × (1-t))} the sector s partial return r _s, the equation (11), the following (12) formula,
r = Σ _s (x _s × r _s ) (12)
Rewrite to Here, x _s is Σ _s x _s = 1 because of the investment weight in the sector s.

Similarly, the benchmark index return R is calculated using the sector benchmark index return R _s by the following equation (13):
R = Σ _s (p _s × R _s ) (13)
Rewrite to Here, p _s is the weight of the sector s in the benchmark index return R, and Σ _s p _s = 1.

Next, using the above equations (12) and (13), the excess return (r−R) is expressed by the following equation (14),
_{r-R = Σ s (x} s -p s) × (R s -R) + Σ s x s × (r s -R s) ... (14)
Deforms as shown in The first term of the equation (14) indicates the “sector tilt effect” in excess return (r−R).

That is, this term is overweighted (x _s > p _s ) for sectors where R _s > R, and underweight (x _s <p _s ) for sectors where R _s <R. For example, it indicates that each excess return (r-R) contributes to be positive.

In other words, the first term in equation (14) is that if the sector benchmark index return R _s has many sectors that are better than the benchmark index return R (excellent profitability of investment), If the sector benchmark index return R _s is lower than the benchmark index return R (the investment profitability is low), it will increase the overall excess return (r−R). Shows that you can.

  The second term of the equation (14) indicates the “brand selection effect” in excess return (r−R).

That is, (r _s -R _s ) in the second term is increased by selecting a good (high investment profitability) stock within the sector s, so a better stock within the same sector s. It shows that the excess return (r-R) can be increased by selecting.

  The above is a method of factoring the excess return into the sector tilt effect and the issue selection effect.

  As a second factor decomposition method, there is a method of factoring excess returns (r-R) into initial holdings and trading during the period.

In this second factor decomposition method of asset management results, a hypothetical portfolio called the initial frozen portfolio is considered. The initial frozen portfolio is an asset when the initial held portfolio is held as it is until the end of the period without any trading. With the return of the case and r _h, r _h can be calculated by equation (10).

Using this r _h , the actual excess return (r−R) is expressed by the following equation (15):
r−R = (r _h −R) + (r−r _h ) (15)
Expand to. The first term in equation (15) represents the excess return due to the initial holdings (or the initial frozen portfolio), and the second term represents the excess return due to trading during the period.

  The above is the factorization method that divides excess returns into initial holding effects and interim trading effects.

2. About the improvement factor decomposition In the improvement factor decomposition, the following preparation is made for the market value V ₀ and V ₁ of the portfolio at the beginning and the end of the period used for the calculation of returns by asset management and the cash flow Σ _t C _t during the period.

From the definitions of V ₀ and V ₁ , the following equations (16) and (17),
V ₀ = Σ _j m _j (0) × p _j (0) (16)
V ₁ = Σ _j m _j (1) × p _j (1) (17)
Can be described as follows. Here, 0 indicates the beginning of the term and 1 indicates the end of the term. m _j (t) indicates the possessed quantity (when t = 0 or t = 1) or the trading quantity (when 0 <t <1) at the time point t. Let p _j (t) be the price per share of brand j at time t.

The above m _j (1) and m _j (0) are expressed by the following equation (18),
m _j (1) = m _j (0) + Σt _(buy) m _j (t) −Σt _(sell) m _j (t) (18)
Have the relationship. Here, Σ _{t (buy)} m _j (t) means the sum of the purchase quantities at all points in time for the issue j. Σ _{t (sell)} m _j (t) means the sum of the sales quantities at all points in time for the issue j. That is, stock j's end-of-period holding quantity m _j (1) is the initial holding quantity m _j (0) plus the purchase quantity Σ _{t (buy)} m _j (t) during the period and the sale quantity Σ _{t ( sell)} m _j (t) can be said to be subtracted.

On the other hand, the cash flow Σ _t C _t during the period is _expressed by the following equation (19),
Σ _t C _t = Σ _{j (buy)} Σ _t m _j (t) × p _j (t)
− Σ _{j (sell)} Σ _t m _j (t) × p _j (t)
− Σ _{j (hold)} m _j (0) × d _j (0 ′)
− Σ _{j (buy)} Σ _t m _j (t) × d _j (t ′)
+ Σ _{j (sell)} Σ _t m _j (t) × d _j (t ′) (19)
Thus, it is possible to break down into trading by brand and interest received or dividend (hereinafter referred to as “interest / dividend”).

Here, the first term of the equation (19) is the sum of the payment amount m _j (t) × p _j (t) when the issue j is purchased at the time t for all purchase points and purchased issues. Is. Since the first term is purchase, it is an increase in assets and the sign is positive.

The second term in equation (19) is the sum of the proceeds m _j (t) x p _j (t) received when the issue j is sold at the time t, for all sale points and sold issues. The sign is negative because it corresponds to a decrease in assets.

The third term of equation (19) is the sum of interest and dividends received during the period for the number m _j (0) of stocks held at the beginning of the period. Note that d _j (0 ′) is the interest / dividend per unit quantity received at time 0 ′ (0 <0 ′ <1) for the issue j held at time 0 (the beginning). The interest income / dividend received will result in a decrease in the asset value due to occurrence, resulting in the outflow of cash from the asset (cash out), so the sign of the third term will be negative.

The fourth term in equation (19) is the sum of interest and dividend received at time t ′ for issues purchased at time t during the period, and d _j (t ′) is the time of issue j purchased at time t. Interest / dividend per unit quantity received at t ′ (t <t ′ <1). Since this term is also cash-out, the sign is negative.

The fifth term of equation (19) is the sum of the interest / dividend deductions for stocks already sold at time t from interest / dividends that should have been received at time t 'during the period. Here, d _j (t ′) is the interest / dividend per unit quantity that can be received at time t ′ (t <t ′ <1) if the issue j sold at time t 1 does not sell. . Since this term is a deduction for interest received and dividends, the sign is positive.

By substituting the equations (16) to (19) prepared as described above into the equation (10) of the return r by the asset-specific modified Dietz method, and calculating the excess return r-R, the following ( 20) Formula,
r−R = w _hold × {Σ _{j (hold)} × _j (0) × r _j ′ (0) −R}
+ W _buy × {Σ _{j (buy)} y _j (*) × r _bj '(*) −R}
−w _sell × {Σ _{j (sell)} z _j (*) × r _sj ′ (*) −R} (20)
Can be obtained.

Here, w _hold indicates the investment weight of the initial holding (V _hold ) in the balance of assets during the period (average value of the balance of assets during the period).
_{w hold = V hold / (V} 0 + Σ t C t × (1-t)) ... (21)
Represented by

w _buy indicates the investment weight of the purchase during the period (V _buy ) in the balance of assets during the period.
_{w buy = V buy / (V} 0 + Σ t C t × (1-t)) ... (22)
Represented by

“w _sell” indicates the investment weight of the sale during the period in the balance of assets during the period.
_{w sell = V sell / (V} 0 + Σ t C t × (1-t)) ... (23)
Represented by

V _hold , V _buy , and V _sell in the above equations (21), (22), and (23) are further expressed by the following equations (24), (25), (26),
V _hold = Σ _{j (hold)} {m _j (0) × [p _j (0) −d _j (0 ′) × (1-0 ′)]} (24)
V _buy = Σ _{j (buy)} Σ _t {m _j (t) × [p _j (t) −d _j (t ′) × (1-t ′) / (1-t)]
× (1-t)} (25)
V _sell = Σ _{j (sell)} Σ _t {m _j (t) × [p _j (t) −d _j (t ′) × (1-t ′) / (1-t)]
× (1-t)} (26)
Respectively represented by

In the above equation (20), x _j (0) indicates the initial investment weight of stock j held at the beginning of the term.
x _j (0) = m _j (0) × {p _j (0) −d _j (0 ′) × (1−0 ′)} / V _hold (27)
Represented by

r _j ′ (0) indicates the modified Dietz method return when the initial holding stock j is held until the end of the period. The following formulas (28), (29),
r _j ′ (t) = r _j (t) / (1−t) (28)
r _j (t) = {p _j (1) −p _j (t) + d _j (t ′)} / {p _j (t) −d _j (t ′) × (1−t ′) / (1− t)} ... (29)
In this case, t = 0 in r _j ′ (t) represented by

From the above equations (28) and (29), the denominator of r _j ′ (t) is {p _j (t) −d _j (t ′) × (1−t ′) / (1−t)} / (1−t), and a value obtained by multiplying this by the quantity m _j (t) corresponds to the average balance of the brand j purchased at time t. This is the V _buy of equation (25), which is calculated for all purchases during the term.

y _j (*) indicates the investment weight of the stock purchased during the period when the portion purchased during the period is regarded as a portfolio. The following formulas (30), (31),
y _j (*) = Σ _{t (buy)} y _j (t) (30)
y _j (t) = {m _j (t) × [p _j (t) −d _j (t ′) × (1-t ′) / (1-t)]
× (1-t)} / V _buy (31)
Represented by In addition, the meaning of the above-mentioned symbol * means the sum of those in the case where a plurality of investment acts are performed for the same issue during the period. That is, since the purchase of the brand j is not limited to once during the period, the weight obtained by totaling the purchases of the brand j during the period is y _j (*) in the equation (30).

r _bj ′ (*) is a weighted average corrected diet method return of purchased stock j during the period,
r _bj ′ (*) = Σ _{t (buy)} ((y _j (t) / y _j (*)) × r _j ′ (t) (32)
Represented by Based on the same idea as the above y _j (*), the return of the brand j, which is a weighted average of the period converted returns r _j ′ (t) of all the brands j purchased during the period with y _j (t) as the weight, R _bj '(*) in the equation (32).

z _j (*) is the investment weight of the issue j during the period when the portion sold during the period is regarded as a portfolio, and is expressed by the following equations (33), (34),
z _j (*) = Σ _{t (sell)} z _j (t) (33)
z _j (t) = {m _j (t) × [p _j (t) −d _j (t ′) × (1-t ′) / (1-t)]
× (1-t)} / V _sell (34)
Represented by

r _sj ′ (*) is the weighted average modified Dietz method return of issue j for sale.
r _sj ′ (*) = Σ _{t (sell)} (z _j (t) / z _j (*)) × r _j ′ (t) (35)
Represented by

  As described above, Expression (20) can be calculated by using Expressions (21) to (35).

  The first term of equation (20) shows the excess return for the initial holding of the asset management, the second term indicates the excess return for the purchase during the term, and the third term indicates the excess return for the sale during the term. It has important features that can be factored into sector tilt effect and brand selection effect.

  In other words, according to equation (20), the initial holding effect, the purchase effect during the period, and the sale effect during the period from the first term to the third term are all returned by the modified Diets method for each asset of equation (12) already explained. It has the same shape as r. For this reason, the initial holding sector tilt effect, the initial holding stock selection effect, and the initial holding stock selection effect by the method similar to the method of the equation (14) in which the excess return (r-R) is decomposed into the sector tilt effect and the stock selection effect. It is possible to obtain the sector tilt effect for the portion purchased during the period, the stock selection effect for the portion purchased during the term, the sector tilt effect for the portion sold during the term, and the issue selection effect for the portion sold during the term.

  A specific factorization method will be described below.

The return of the portfolio part in each excess return in equation (20) is the sum of the product of the weight by issue and the return by issue, and is in the form of Σ _j x _j × r _j. Is transformed into sectoral returns.

Here, the sum of x _j for sector s is defined as x _s, and the sector return calculated by weighted average of r _j with x _j is denoted as r _s , the following equations (36), (37),
x _s = Σ _j x _j (36)
r _s = Σ _j ((x _j / x _s ) × r _j ) (37)
become that way. Here, Σ _j indicates the sum for all issues j belonging to the sector s.

At this time, Σ _j x _j × r _j is the following equation (38):
Σ _j x _j × r _j = Σ _s x _s × Σ _j ((x _j / x _s ) × r _j )
_{= Σ s x s × r s} ... (38)
It can be deformed as follows.

By this method, the first to third terms of equation (20)
Initial holding effect = w _hold × {Σ _{j (hold)} x _j (0) × r _j '(0) −R} (1)
Purchase effect during the period = w _buy × {Σ _{j (buy)} y _j (*) × r _bj '(*) -R} (2)
Sale effect during period = w _sell × {Σ _{j (sell)} z _j (*) × r _sj '(*) −R} (3)
, The above formulas (1), (2), (3) are replaced by the following formulas (4)-(9),
Initial sector sector tilt effect = w _hold × {Σ _s (x _{s (hold)} −p _s ) × (R _s −R)} (4)
Stock selection effect at the beginning of the fiscal period = w _hold × {Σ _s × _{s (hold)} × (rs _(hold) −R _s )} (5)
Sector tilt effect for purchases during the period = w _buy x {Σ _s (x _{s (buy)} -p _s ) x (R _s -R)} (6)
Stock selection effect during the period = w _buy × {Σ _s × _{s (buy)} × ( _{rs (buy)} −R _s )} (7)
Sector tilt effect for sale during period = w _sell × {Σ _s (x _{s (sell)} -p _s ) × (R _s -R)} (8)
Stock selection effect during the period = w _sell × {Σ _s x _{s (sell)} × ( _{rs (sell)} −R _s )} (9)
The factor can be decomposed.

Here, x _{s (hold)} = Σ _{j, s (hold)} x _j (0), where j, s (hold) is the stock j in the initial holding of sector s. is there.

R _s is called a sector benchmark index return and is a benchmark index return for sector s.

R = Σ _s (p _s × R _s ), where p _s is the weight of sector s in the benchmark.

Also, r _{s (hold)} = Σ _{j, s (hold)} ((x _{j (hold)} / x _{s (hold)} ) × r _j ′ (0)).

Further, it can be expressed as x _{s (buy)} = Σ _{j, s (buy)} y _j (*). In the equation, j, s (buy) is the stock j in the purchases during the period of sector s.

Also, r _{s (buy)} = Σ _{j, s (buy)} ((y _j (*) / x _{s (buy)} ) × r _bj ′ (*)).

Further, it can be expressed as x _{s (sell)} = Σ _{j, s (sell)} z _j (*). In the equation, j, s (sell) is the stock j in the sale of sector s during the period.

Also, r _{s (sell)} = Σ _{j, s (sell)} ((z _j (*) / x _{s (sell)} ) × r _sj ′ (*)).

  The above is the improvement factor decomposition. In order to help understand the improvement factor decomposition, FIG. 6 shows examples of various cases in which initial holding, purchasing during the period, and selling during the period occur. For example, according to Case 1, since stocks were held at the end of October, there was an initial holding in November and no action was taken in November. In addition, stocks from October continue to be held at the end of November, so initial holdings will also occur in December. Also, for example, according to Case 4, since the stock was held at the end of October, it was held for the first time in November, and in November the stock was purchased and sold. There are both purchases and sales during the fiscal year, and as a result of the sale, stocks ceased at the end of November. Also, for example, according to Case 7, since no shares were held at the end of October, there was no initial holding in November, and in November the shares were purchased but not sold. Although it was purchased during the term, there was no sale during the term, and as a result, since the stock was held as of the end of November, the initial holding will occur in December.

  As an asset management result analysis support system for providing such information to the user (for example, a person who analyzes the result of asset management) by performing such improvement factor decomposition and for example, as illustrated in FIG. System is conceivable.

  FIG. 1 shows a configuration example of an asset management result analysis support system according to the present embodiment.

  A user terminal (for example, a computer such as a personal computer) 300 which is a terminal used by a user and a server 313 for supporting the asset management result analysis for the user are connected to be communicable via a communication network 351. ing. As the communication network 351, for example, the Internet, a LAN (Local Area Network), a SAN (Storage Area Network), or the like can be adopted.

  The server 313 includes, for example, a plurality of databases (hereinafter referred to as DBs) 361, 363, and 366, and a controller 302.

  Each DB can be constructed by, for example, a drive (hereinafter simply referred to as “disk drive”) of one or more disk-type storage media (for example, a hard disk or a DVD (Digital Versatile Disk)). At least one of the plurality of DBs 361, 363, and 366 may be provided outside the server 313 and connected to the server 313 so as to be communicable. In that case, the DB can be a database server.

  The DB 361 is a DB (hereinafter referred to as “change / balance DB”) that stores a history of asset changes for asset management. The change history can be composed of, for example, change data representing the contents and results of each change, and end-of-period balance information indicating information relating to the balance at the end of each end (for example, at the end of the month).

The transfer / balance DB 361 stores a large number of transfer data. Each change data may be stored separately for each period (for example, for each month). FIG. 5A shows each transfer data related to asset management from November to December 2005, for example. Each change data includes, for example, the date when the change was made (change date), the type of change (for example, the balance at the end of the month, purchase, sale, or interest / dividend), and the transferred issue (for example, issue code). , The contents of the change (for example, the quantity of the stock purchased or sold, the unit price of the stock, and the amount obtained by multiplying the quantity and the unit price). Each transfer data includes a book value balance, a time-weighted amount (this is calculated and recorded when the transfer data is stored, for example), a month-end market price, and a month-end remaining evaluation amount. At least one may be included. The transfer data may include basic information for each brand corresponding to each of V _hold and V _buy . Here, the brand-basis basic information of V _hold is a value in the outermost parenthesis of the equation (24). The basic information for each brand of V _buy is a value after Σ _{t in} equation (25).

  The DB 363 is a DB storing evaluation market price data (hereinafter, evaluation market price DB). For example, as illustrated in FIG. 5B, the evaluation market price data includes a brand (for example, brand code), a date (year / month / day), and a market price for evaluation per share of the brand on that date (for example, closing price). It is included. By searching the evaluation market price DB 363 using the brand and date as a search key, it is possible to acquire the market price for evaluation of that brand on that date. The DB 363 can also store R (benchmark index return) for each predetermined period.

  The DB 366 is a DB for storing calculation result data (hereinafter, calculation result DB). The calculation result data is data representing a result calculated based on a large number of change data belonging to a specified period by a method described later. The calculation result data includes, for example, a specified period and a calculation result. The calculation result is, for example, at least one of an initial holding effect, an interim purchase effect, an interim sale effect, and an overall rate of return.

  The controller 302 is a device for controlling the operation of the server 313. The controller 302 includes a network I / F 307 that is an interface device (hereinafter referred to as I / F) for a communication network, a disk I / F 305 that is an I / F for each disk drive constructing a DB, a memory 311, And a processor 309. The memory 311 is one or a plurality of volatile and / or nonvolatile memories. The memory 311 includes, for example, a work area for the processor 309 and an area for storing a computer program that is read and executed by the processor 309. An example of the computer program is a control program 1.

  FIG. 2 shows a configuration example of the control program 1.

  The control program 1 can be, for example, an application program that operates on an operating system (not shown) of the server 313. This control program 1 is a program that can perform the above-described improvement factor decomposition, but can not only simply perform the improvement factor decomposition, but also calculate the effect of sale during the period without the interest / dividend data, Therefore, it is devised so that the consumption of storage resources can be suppressed and the cost can be greatly saved.

  Specifically, for example, the control program 1 includes a data reading unit 601, an overall profit rate calculation unit 30, a determination unit 602, an initial holding effect calculation unit 2a, an interim purchase effect calculation unit 2b, and an interim sale effect. And a calculator 2c.

  The data reading unit 601 can read from the transfer / balance DB 361 M transfer data that is smaller than N transfer data related to the calculation target period. The value of M may be a fixed value or a variable value. The data reading unit 601 can read N pieces of change data by reading M pieces of change data one by one.

The total rate of return calculation unit 30 updates the types of values necessary for calculating the total rate of return (the above-mentioned return r) using the predetermined types of values recorded in the transfer data sequentially read out. This is a subprogram that can calculate the overall rate of return using the updated value. Specifically, for example, the total rate of return calculation unit 30 updates the values of Σ _t C _t and Σ _t C _t × (1−t) in the above equation (10), and uses these updated values. Thus, the overall rate of return can be calculated by equation (10).

  Using the change type in the read change data, the determination unit 602 determines which of the plurality of calculation units 2a and 2b is to be a distribution destination of the transfer data, and determines the determined distribution Transfer data can be sorted first.

The initial holding effect calculation unit 2a is a subprogram capable of calculating the initial holding effect using the distributed change data. The initial holding effect calculation unit 2a, for example, for each brand,
・ Calculate the period return r _j ′ (0) by stock using the value of the stock price p _j (1) at the end of the period, the initial stock price p _j (0) and the interest and dividend d _j (0 ′) at the beginning of the period
· Beginning quantity m _j (0), using the beginning of the year stock price p _j (0) and the beginning of the interest and dividend amounts d _j (0 '), to calculate the mark-to-market capitalization V _hold of the beginning portion, of the V _hold and assets Calculating the investment weight (contribution weight) w _hold at the beginning of the term using the total balance ((V ₀ + Σ _t C _t × (1−t));
・ Calculate the weight x _j (0) by issue using the initial quantity m _j (0), initial share price p _j (0), initial interest / dividend d _j (0 '), and V _hold calculated above. To do,
Calculate the initial holding effect using the calculated r _j '(0) and w _hold , x _j (0) and the benchmark index return R;
Can be executed.

The interim purchase effect calculation unit 2b is a subprogram that can calculate the interim purchase effect using the distributed change data. The purchase effect calculation unit 2b during the period, for example, for each brand, the following processing,
・ Calculate the market capitalization value V _buy of the purchase part during the period using the purchase quantity m _j (t) during the period, the stock price p _j (t) at the time of purchase during the period and the interest / dividend d _j (t ') during the period And calculating the investment weight (contribution weight) w _buy of the portion purchased during the period using V _buy and the total balance of assets.
・ Use m _j (t), p _j (t), d _j (t ') and V _buy to calculate the weight y _j (t) for each transfer, and use y _j (t) and V _buy , Calculating weighted weight y _j (*) for each issue,
・ By using the value of stock price p _j (1) at the end of the period, p _j (t) and d _j (t '), the period-specific return r _j ' (t) is calculated, and r _j '(t), y _j (T) and y _j (*) are used to calculate the weighted return r _bj '(*) for each issue,
・ Use the calculated w _buy , y _j (*) and r _bj ′ (*) and the benchmark index return R to calculate the purchase effect during the period,
Can be executed.

  The interim sale effect calculation unit 2c is a subprogram that can calculate the interim sale effect by subtracting the sum of the calculated initial holding effect and the interim purchase effect from the calculated overall rate of return.

  The above is the description of the server 313 according to the present embodiment.

By the way, the controller 302 can update the above-described transfer / balance DB 361 each time a change is made. For example, when a change occurs for a certain issue, change data for that issue is added to the change / balance DB 361. The balance information of the transfer data can be generated based on the balance information of the transfer data that already exists for the brand (for example, transfer data corresponding to the previous transfer) and the content of the current transfer. More specifically, for example, each time each change is made, the brand-specific basic information corresponding to the transferred brand (brand-basic information corresponding to each of V _hold and V _buy ) can be updated. Hereinafter, for a certain stock, the length of the calculation target period (for example, a certain month) is set to 1, and the number of shares at the beginning of the calculation target period (that is, the number of shares held immediately before the calculation target period) is Q. A specific example will be described as (Q is an integer of 1 or more).

(A) Update V _hold basic information.

When the type of transfer is “interest / dividend”, the controller 302 calculates (1-0 ′) and d _j from the transfer date (the date on which interest / dividend has entered) and the interest / dividend in the calculation target period 1. (0 ') can be obtained. Further, it can be determined from the balance information of the period immediately before the calculation target period that m _j (0) = Q. Further, p _j (0) can be obtained from the evaluation market price DB 363 as the closing price of the day immediately before the calculation target period. Therefore, using equation (24), the basic information on V _hold is obtained, the basic information on V _hold up to that time is updated to the obtained basic information, and the updated basic information is used as the current transfer data. Can be recorded.

(B) Update of V _buy basic information When the type of transfer is “Purchase”, the controller 302 calculates m _j (t), p _j (t) and (1) from the current transfer date, the number of shares purchased and the purchase price. -T) can be obtained. In addition, the controller 302 can check whether or not interest / payout is performed for the purchased brand during the calculation target period. For example, the controller 302 checks whether or not there is change data whose type of change is “interest / payout” among a plurality of change data belonging to the calculation target period. As a result, (1-t ′) and d _j (t ′) can be obtained from the transfer date and interest / dividend, if any. The controller 302 uses the basic information of V _buy , m _j (t), p _j (t), and (1-t) up to immediately before the current change, using the equation (25) ((1-t ′) And d _j (t ′) if they are also obtained), the basic information of V _buy after this change can be obtained.

As described above, when adding the transfer data, the provisional basic information for each brand of V _hold and V _buy in the balance information can be updated. When the calculation period has passed, the basic information of the V _hold and V _buy in the final of change data for each security in the calculation period, stock-specific basic information V _hold and V _buy in the calculation period Can be determined respectively. In addition, the controller 302 can include the basic information for each brand of V _hold and V _buy in the final transfer data in the end-of-period balance information.

  Now, a flow of processing performed by the control program 1 devised so that interest / dividend data can be made unnecessary for calculation of the sale effect during the period will be described.

  FIG. 3 shows an example of the flow of processing performed by the processor 309 that reads and executes the control program 1.

  The processor 309 that executes the control program 1 (hereinafter referred to as “control program 1” for the sake of convenience), for example, performs the processing of FIG. 3 when receiving a request from the user terminal 300 for analysis of asset management results. Can start.

  The control program 1 reads transfer data relating to the calculation target period from the transfer / balance DB 361 (step S1). Specifically, for example, the data reading unit 601 of the control program 1 transfers the change data (change type is the transfer type) corresponding to the period immediately before the calculation target period (for example, the period specified by the user) from the change / balance DB 361. Read the transfer data (the balance at the end of the month). Alternatively, the data reading unit 601 of the control program 1 reads one or more pieces of change data among a large number of pieces of change data within the calculation target period, and accumulates them in the memory 311.

Next, the overall rate of return calculation unit 30 of the control program 1 uses a predetermined type of value (for example, at least the change date and change content among the transfer date, transfer content, and time-weighted amount) included in the transfer data. Then, the values of Σ _t C _t and Σ _t C _t × (1−t) in the above equation (10) are updated on the memory 311 (S2).

  Next, the control program 1 determines the type of change (end-of-month balance, buy or sell) of the read change data (S3), and determines whether to perform S4A or S4B according to the determination result. Specifically, for example, when the determination unit 602 of the control program 1 reads the first change data in FIG. 5A, it determines that the change type is the end-of-month balance. By allocating to the holding effect calculation unit 2a, S4A (update of the value related to the initial holding effect) can be performed. Further, when the determination unit 602 reads the second change data in FIG. 5A, the determination unit 602 determines that the change type is buy, and in that case, distributes the change data to the interim purchase effect calculation unit 2b, S4B (update of the value related to the purchase effect during the period) can be performed. Furthermore, when the third change data in FIG. 5A is read, the determination unit 602 determines that the change type is sell, and in that case, the change data can be discarded. This is because even if there is no transfer data, the effect of sale during the period can be calculated as described later. The control program 1 can delete the transfer data stored in the memory 311 from the memory 311 after updating the necessary types of values on the memory 311 using the read transfer data.

  The control program 1 determines whether or not all change data related to the calculation target period has been read (S5), and if not read (NO in S5), S1 is performed again. The control program 1 repeats the above process every time the change data is read.

  When the control program 1 determines that all the transfer data related to the calculation target period has been read (YES in S5), the overall return rate calculation unit 30 calculates the overall return rate using the updated values. The initial holding effect calculation unit 2a calculates the initial holding effect, and the interim purchase effect calculation unit 2b calculates the interim purchase effect (S6). Then, the interim sales effect calculation unit 2c calculates the interim sales effect by subtracting the sum of the calculated initial holding effect and the interim purchase effect from the calculated overall rate of return (S7).

  Thereafter, the control program 1 can write the various effects and the overall profit rate calculated in the calculation result DB 366 (S8) or print them (S9).

  The above is an outline of the flow of processing that can be performed by the control program 1.

  Now, the calculation of the overall rate of return, the initial holding effect, the purchasing effect during the period, and the selling effect during the period will be described in detail. In the following, specific examples are used as appropriate, but in that case, FIGS. 5A to 5B are referred to. Therefore, the calculation target period is November 2005, and the immediately preceding period is October 2005. Therefore, in November 2005, it was assumed that only the stock code 1234 was traded, and all the stocks of the stock code 1234 had been sold by the end of October just before that, and the stocks of the stock code 1234 were also held. (In short, the situation is that no stock of any brand is held).

  (A) Calculation of overall rate of return (S1, S2, S5, S6).

The total profit rate calculation unit 30 uses the predetermined type of value (for example, at least the change date and change content among the change date, change content, and time-weighted amount) included in the change data, the above equation (10) The values of Σ _t C _t and Σ _t C _t × (1−t) are updated on the memory 311. Specifically, for example, when the first transfer data in FIG. 5A is read, the total profit rate calculation unit 30 can write V ₀ = 0 on the memory 311 (the end of October). Because the end-of-month evaluation of the balance is 0) When the second transfer data is read, 2,000,000 is written as Σ _t C _t (since C _t of the first case is 2,000,000), and 1,266,667 is set as Σ _t C _t × (1-t). Can be written (because the time weighted amount is that amount). If the third change data is read, write 900,000 (= 2,000,000-1,100,000) as sigma _t C _t, and, as _{Σ t C t × (1-} t) 1,083,334 (= 1,266,667-183,333) to write Can do. When the fourth transfer data is read, V ₁ = 1,120,000 can be written (because the end-of-month evaluation of the end-of-month balance in November is 1,120,000).

When the 1st to 4th change data is read, it means that the transfer data relating to the calculation target period has been read, so that the total profit rate can be calculated in S6. Specifically, according to the above examples and _{(10), V 1 -V 0 -Σ t C} t = 220,000 _{becomes, V 0 + Σ t C t} × (1-t) = 1,083,334 , and the thus, r ≒ 0.2 (about 20%). For example, it may be determined whether or not the read change data is the last change data related to the calculation target period in S2, and r is calculated when it is determined as the last change data.

  (B) Calculation of initial holding effect (S1, S2, S3, S4A, S5, S6).

  For example, when the control program 1 reads balance data corresponding to the period immediately before the calculation target period (in FIG. 5A, change data of the change date “2005/10/31” and change type “end of month balance”). The affiliation is determined to be held at the beginning of the term, and S4A is performed. FIG. 4A shows a typical formula necessary for calculating the initial holding effect.

The control program 1 acquires the brand, the number of shares held _{j j} (0) and the stock price p _j (0) from the read balance data, and acquires the brand, the number of shares held _{j j} (0) and the stock price p _j (0 ) Is written in the memory 311. According to the example of FIG. 6A, it is written that the brand code 1234, m _j (0) = 0, and p _j (0) = 950.

Further, if there is interest / dividend change data having the stock code, the control program 1 obtains (1-0 ′) from the change date (that is, interest / dividend date) in the change data, and the memory 311 And the unit price (that is, interest / dividend) d _j (0 ′) is acquired from the contents of the change in the change data and written in the memory 311. According to the example of FIG. 5A, since the interest / dividend of the stock code 1234 is not within the calculation target period, the interest / dividend date and the interest dividend are not written.

Further, the control program 1 acquires the stock price p _j (1) at the end of November 2005 using the brand code and change date of the acquired balance data as a search key, and writes them in the memory 311. According to the example of FIG. 5A, since the end-of-month market price recorded in the evaluation market price DB 363 is also recorded as transfer data, there is no need to refer to the evaluation market price DB 363, and the stock price p _j (1) of the stock code 1234. 1120 is recorded in the memory 311.

The control program 1 stores m _j (0) (= 0), p _j (0) (= 950), (1-0 ′) (= none), d _j (0 ′) (= None) and p _j (1) (= 1120) are used to calculate r _j ′ (0) according to equations (28) and (29), and the calculated r _j ′ (0) is written in the memory 311. . Specifically, 170 / 950≈0.18 is calculated, and r _j ′ (0) = 0.18 is written in the memory 311.

Furthermore, the control program 1 uses the expression m (24) to _hold V _hold using m _j (0), p _j (0), (1-0 ′) and d _j (0 ′) written in the memory 311. of basic information (hereinafter, V _hold basic information) asked for, the V _hold basic information, it can also be written to the memory 311. Instead, the control program 1 acquires V _hold basic information (for example, information written in the last transfer data in this period) from the transfer / balance DB 361 corresponding to the calculation target period, and acquires the acquired V _hold Basic information may be written in the memory 311. Here, for example, for the stock code 1234, since m _j (0) = 0 and p _j (0) = 950, the V _hold basic information is 0. However, as another example, m _j (0 ) = 1,000, 1,000 × 950, and V _hold basic information is 950,000.

The control program 1 acquires m _j (0), p _j (0), V _hold basic information, r _j ′ (0), etc. for other brands by repeating S1 to S5. In that case, the control program 1 can write the information in the memory 311 for each brand. The control program 1, when it becomes YES in S5, the (16) equation, to calculate the V ₀ (or, based on the number of retained and stock by each stock in the closing balance information, obtains the V ₀₎ The obtained V ₀ can be written in the memory 311. If YES in S5, the control program 1 adds V _hold basic information of all brands written in the memory 311 according to the equation (24), calculates V _hold, and _stores it in the memory. 311 can be written. Here, according to the example of FIG. 5A, since m _j (0) = 0 and no stock is held at the end of October, both V ₀ and V _hold are 0. If it is held at the end of October, neither V ₀ nor V _hold will be 0.

The control program 311 can obtain x _j (0) for each brand by the equation (27) using the V _hold basic information and V _hold for each brand and write it in the memory 311. Here, V _hold basic information and V _hold of the stock code 1234 are 0, and since no stock is held at the end of October, x _j (0) is also 0, but for example, m _j If (0) = 1,000, V _hold basic information is 950,000 as described above, and if only stock code 1234 is held, V _hold is also 950,000, so x _j (0) is 950,000 / 950,000 = 1

S1~S5 is, when repeated until YES in S5, the asset that has been updated in S2 Sotairazan _{(V 0 + Σ t C t} × (1-t)) is recorded in the memory 311 . The control program 1 uses the total balance of the updated assets written in the memory 311 and the V _hold and calculates w _hold by the equation (21), and writes the calculated w _hold in the memory 311. be able to. Specifically, the total balance of assets is 1,083,334 as described above, and V _hold is 0, so w _hold is 0 / 1,083,334 = 0. For example, if V _hold is 950,000 as in the above example, 950,000 / 1,083,334≈0.877 and w _hold = 0.877 is written in the memory 311.

The control program 1 can calculate the initial holding effect by the equation (1) using w _hold , x _j (0), r _j ′ (0) and R written in the memory 311 (Note that R (For example, R = 4.000) can be read from a predetermined storage area (for example, evaluation market price DB 363) at an arbitrary timing and stored in the memory 311). Specifically, for example, if w _hold = 0, the initial holding effect is also 0, but as another example, if w _hold = 0.877, x _j (0) = 1 and r _j ' Since (0) is 0.18, the initial holding effect is 0.877 × (1 × 0.18−4.000) = about −3.350.

  (C) Calculation of purchase effect during the period (S1, S2, S3, S4B, S5, S6).

  When the control program 1 reads out the change data in which the change type is purchase (in FIG. 5A, the change date “2005/11/11” / change data of the change type “buy”), the affiliation is an interim purchase. And S4B is performed. FIG. 4B shows a typical formula necessary for calculating the purchase effect during the period.

The control program 1 acquires the brand, the number of purchased shares m _j (t) and the purchase price p _j (t) from the read change data, and acquires the acquired brand, the number of owned shares m _j (t), the stock price p _j (t ) And (1-t) based on the change date are written in the memory 311. According to the example of FIG. 5A, it is written that the brand code 1234, m _j (t) = 2,000, p _j (t) = 1,000, (1−t) ≈0.633.

Further, if there is interest / dividend change data having the stock code, the control program 1 obtains (1-t ′) from the change date (that is, interest / dividend date) in the change data, and the memory 311. And the unit price (that is, interest dividend) d _j (t ′) is acquired from the contents of the change in the change data and written in the memory 311. According to the example of FIG. 5A, since the interest / dividend of the stock code 1234 is not within the calculation target period, the interest / dividend date and the interest dividend are not written.

Further, the control program 1 acquires the stock price p _j (1) at the end of November 2005 using the stock code and the change date of the acquired change data as a search key, and writes them in the memory 311 (by the process of S4A) If you have already done this, you don't have to do this). The stock price p _j (1) of the stock code 1234 is 1120.

The control program 1 includes m _j (t) (= 2,000), p _j (t) (= 1,000), (1-t) (= 0.633), (1-t ′) (= none) written in the memory 311. ), D _j (t ′) (= none) and p _j (1) (= 1120) are used to calculate r _j ′ (t) according to equations (28) and (29), and the calculated r _j ′ (t) is written in the memory 311. Since r _j (t) = 120/1000 = 0.12, r _j ′ (t) = 0.12 / 0.633≈0.190.

Further, the control program 1 stores m _j (t) (= 2,000), p _j (t) (= 1,000), (1-t) (= 0.633), (1-t ′) ( = no) and using d _j (t ') (none =), (a 25), and updates the provisional basic information V _buy, the provisional basic information V _buy the updated memory You can also write 311. This information is determined as V _buy basic information when all the transfer data for one brand is read out. Further, when YES is obtained in S5, V _buy itself can be calculated and written in the memory 311 by the equation (25). The control program 1 acquires the V _buy basic information (for example, information written in the final change data in this period) for each issue corresponding to the calculation target period from the transfer / balance DB 361 and acquires it. V _buy basic information may be written in the memory 311. Here, the V _buy basic information of the stock code 1234 is 2,000 × 1,000 × 0.633 = 1,266,000. Since no shares other than the stock code 1234 are held, V _buy is 1,266,000.

The control program 311 can obtain y _j (t) for each brand by the equation (31) using the V _buy basic information and V _buy for each brand and write them in the memory 311. Further, the control program 311 can obtain y _j (*) by using equation (30) using y _j (t) of each brand and write it in the memory 311. Since V _buy basic information is 1,266,000 as described above and only the stock code 1234 is held, V _buy is 1,266,000, so y _j (t) of the stock code 1234 is 1,266,000 / 1,266,000 = 1. _j (*) is also 1.

S1~S5 is, when repeated until YES in S5, the asset that has been updated in S2 Sotairazan _{(V 0 + Σ t C t} × (1-t)) is recorded in the memory 311 . The control program 1, by using the total average balance and V _buy assets after updating written in the memory 311, the (22) equation, to calculate the w _buy, write the calculated w _buy the memory 311 be able to. Here, according to the above-mentioned example, since the total balance of assets is 1,083,334 and V _buy is 1,266,000, w _buy is 1,266,000 / 1,083,334≈1.169.

Further, the control program 1 uses y _j (*) (= 1), r _j ′ (t) (= 0.190) and y _j (t) (= 1) of each brand written in the memory 311. R _bj ′ (*) can be calculated from the equation (32) and written in the memory 311. r _bj '(*) is 0.190.

The control program 1 uses w _buy (= 1.169), y _j (*) (= 1), r _bj '(*) (= 0.190), R (= 4.000) written in the memory 311 (2) The purchase effect during the period can be calculated and written in the memory 311 in S6 according to the equation. According to the above example, the purchase effect during the period is 1.169 × (1 × 0.190−4.000) = about −4.454.

  (D) Calculation of the effect of sale during the period (S7).

  As described above, the interim sale effect calculation unit 2c calculates the interim sale effect by subtracting the sum of the calculated initial holding effect and the interim purchase effect from the calculated overall rate of return. In the above example, r≈0.2, initial holding effect = 0, and interim purchase effect≈−4.454, so interim sales effect = 0.2− (0 + −4.454) = 4.654. In another example, since r≈0.2, initial holding effect≈−3.350, and interim purchase effect≈−4.454, the interim sales effect = 0.2 − (− 3.350 + −4.454) = 8.004.

  The above is the description of the present embodiment.

  According to the above-described embodiment, unlike the conventional factor decomposition in which only the effect of buying and selling during the period can be obtained in addition to the initial holding effect, it is possible to execute the improvement factor decomposition that further divides the effect during the period into purchase and sale it can.

  In this embodiment, the sale effect during the period can be calculated without the interest / dividend data of the sold brand.

  In other words, the improvement factor decomposition can calculate the initial holding effect, the purchasing effect during the period, and the selling effect during the period, but in calculating various effects, if there is interest / dividend within the calculation target period, The interest / dividend data (data including the interest / dividend date and interest / dividend) of the purchased or sold stocks that are sold is required.

  If interest / dividends are generated for stocks held before the calculation target period, or for stocks that have been purchased by the purchase during the calculation target period, the control program 1 uses the interest / dividend data. Can be acquired from the transfer / balance DB 361. This is because, if interest / dividend is generated for the stocks held, transfer data including interest / dividend date and interest / dividend will be added to the transfer / balance DB 361 as shown in FIG. 5A. is there.

  However, since an interest / dividend does not occur for a stock that is no longer held by selling, the interest / dividend data of the stock is not stored in the transfer / balance DB 361. Cannot obtain interest / dividend data on sold stocks. Therefore, when calculating the effect of sale during the period, if interest / dividend on the sold stock is incurred on the day of the calculation period, the interest / dividend of the sold stock is purposely calculated only for the purpose of calculating the rate of return. There is a cost for purchasing data from an external vendor or the like and managing the data so that it can be used systematically.

  According to the above-described embodiment, first, the overall return rate, the initial holding effect and the interim purchase effect are calculated, and then the overall return rate-(the initial holding effect + the interim purchase effect) is calculated, thereby calculating the interim sales effect. The control program 1 and the processing flow are devised so that they can be obtained. As a result, it is possible to save the labor and cost of purchasing, processing, and accumulating the interest / dividend data of the brands that are no longer held by the sale only for calculating the profit rate.

  The embodiment of the present invention has been described above, but this is an example for explaining the present invention, and the scope of the present invention is not limited to this embodiment. The present invention can be implemented in various other forms.

  For example, in the loop of S1 to S5, the control program 1 can continuously read the change data of a certain brand and then continuously read the change data of another brand. Specifically, for example, the control program 1 can also send a SQL (Structured Query Language) representing a request for acquisition of transfer data belonging to a certain brand to the transfer / balance DB 361. In the same manner, the control program 1 continuously reads the change data of the issues belonging to the same sector, reads all the change data belonging to the sector, and then continuously reads the change data of the issues belonging to another sector. I can read it.

  Further, for example, in the loop of S1 to S5, the control program 1 first reads the interest / dividend change data (fund interest / dividend) belonging to the calculation target period, and thereafter, the other change You may read the data.

The structural example of the asset management result analysis support system which concerns on one Embodiment of this invention is shown. The structural example of the control program 1 is shown. An example of the flow of processing performed by the processor 309 that reads and executes the control program 1 is shown. FIG. 4A shows a typical formula necessary for calculating the initial holding effect. FIG. 4B shows a typical formula necessary for calculating the purchase effect during the period. FIG. 5A shows an example of each change data stored in the change / balance DB. FIG. 5B shows an example of the market price data for evaluation that constitutes the market price DB 363 for evaluation. The figure for demonstrating the case where the initial holding, the purchase during the term, and the sale during the term each occur.

Explanation of symbols

DESCRIPTION OF SYMBOLS 1 ... Control program 2a ... Initial holding effect calculation part 2b ... Interim purchase effect calculation part 2c ... Interim sale effect calculation part 309 ... Processor 311 ... Memory 361 ... Transfer and balance DB 363 ... Evaluation market price DB w _hold ... Investment weight w _buy at the beginning of the term w _buy … Investment weight for the purchase during the term occupying the asset balance during the term w _sell … Investment weight for the sale during the term occupying the asset balance during the term x _j (0)… j's initial investment weight r _j '(0) ... Revised Diets method return when stock j is held until the end of the period R ... Benchmark index return y _j (*) ... When the portion purchased during the period is regarded as a portfolio the investment weight r _bj '(*) ... purchase brand j of the period in the weighted average of the modified Dietz method return z _j (*) ... during the period the sale portion of the period during the purchase brand j Investment weight r _sj '(*) ... our initial investment weight R of beginning assets of each period during the weighted average of the modified Dietz method return x _{s (hold)} ... sector of sale brand j interim sale brand j at the time was regarded as over portfolio _s … sector benchmark index return p _s … weight of sector s in benchmark x _{s (buy)} … investment weight of asset purchased in sector s x _{s (sell)} … investment weight of asset sold in sector s r _{s (Hold)} … Adjusted Diets method return based on assets initially held in sector s r _{s (buy)} … _Adjusted Diets method return based on assets purchased in sector s r _{s (sell)} … _Adjusted Diets method return based on assets sold in sector s during period

Claims (5)

An interface device for a storage device storing a plurality of transfer data relating to inflow or outflow of assets ;
Memory,
A processor connected to the interface device and the memory;
Have
Each change data includes (a) the transfer date that is the date when the asset is changed, (b) the transfer type that indicates either balance, outflow, inflow, or interest / dividend, and (c) the transfer issue that is the transferred issue. And (d) including the amount of money transferred,
The processor is
(A) Two or more change data related to a calculation target period, which is specified based on a change date and change type included in each of the plurality of change data, is read from the storage device ,
(B) read more changes data change date, each containing a using change type and change amount, calculates the overall rate of return on the memory,
(C) Based on the change type of each read change data , whether each read change data is used for calculating the initial holding effect, calculating the purchase effect during the period, or calculating the effect of selling during the period discriminated,
(D) Of the two or more read change data, using only the transfer data used for calculating the initial holding effect and the transfer data used for calculating the purchase effect during the period, the asset management results are shown as the initial holding effect. and factor analysis to the three types of effect of, period during the purchase effect and interim sale effect,
(E) outputs information based on the factor analysis results,
Wherein the processor is in the (D), the possession by a interest and dividend data as change data of which or stocks purchased interest and dividend date the calculated period is a change date its interest and dividend data includes If it belongs, the initial holding effect and the purchasing effect during the period are calculated using the determination result and the interest / dividend data of the stock, and then the initial holding effect and the purchasing effect during the period are calculated from the calculated overall rate of return. The effect of selling during the period is calculated on the memory by subtracting
Asset management result analysis support system. In the (B), the processor updates the total balance of assets in the calculation target period on the memory based on each read change data,
In the (D), the processor performs the factorization using the total balance of assets prepared on the memory .
The asset management result analysis support system according to claim 1. The processor writes the read change data to the memory each time the change data is read,
Wherein the processor is using the change data written into said memory, said is Ru updates performed at the decomposition analysis of the (D) each time the ends, deletes the change data used for the update from the memory ,
The asset management result analysis support system according to claim 1. (A) the date of transfer, which is the date on which the asset was transferred, (b) the transfer type representing either balance, outflow, inflow, or interest / dividend, (c) the transferred issue that is the transferred issue, and (d ) changes and plurality store different dynamic data about the inflow or outflow of assets include a change amount is the amount the storage instrumentation placed al, computer, based on the change date and change the type of each of the plurality of change data includes Reading two or more transfer data related to the calculation target period specified in
A computer calculating an overall rate of return on a memory using a transfer date, a transfer type and a transfer amount included in each of the two or more read transfer data;
Based on the type of change of each read change data , the computer determines whether each read change data is used for the calculation of initial holding, calculation of purchase during the period, or calculation of effect of sale during the period. Steps,
Of the two or more read change data, the computer uses only the transfer data used to calculate the initial holding effect and the transfer data used to calculate the purchase effect during the period, and the results of the asset management are shown as the initial holding effect. A step of factoring into three types of effects: purchase effect during the period and sale effect during the period;
A computer outputting information based on the result of the factorization,
Wherein the factor decomposing, if interest and dividend date is interest and a payout data transfers date the interest and dividend data includes as change data stocks were or purchased it holds belongs to the computation period, The initial holding effect and the purchasing effect during the period are calculated using the determination result and the interest / dividend data of the stock, and then the initial holding effect and the purchasing effect during the period are calculated from the calculated overall rate of return. Calculate the effect of selling during the period on the memory by subtracting,
Asset management performance analysis support method. (A) the date of transfer, which is the date on which the asset was transferred, (b) the transfer type representing either balance, outflow, inflow, or interest / dividend, (c) the transferred issue that is the transferred issue, and (d ) changes and plurality store different dynamic data about the inflow or outflow of assets include a change amount is the amount the storage instrumentation placed al, is specified based on change date and change the type each comprising a plurality of change data Reading two or more transfer data related to the calculation target period ;
Calculating an overall rate of return on the memory using the change date, change type and transfer amount included in each of the two or more read change data;
Determining whether each read change data is used for calculation of initial holding, calculation of purchase effect during period, and calculation of sale effect during period based on the change type of each read change data ; ,
Of the two or more read transfer data, only the transfer data used to calculate the initial holding effect and the transfer data used to calculate the purchase effect during the period are used to obtain the results of asset management as the initial holding effect and purchase during the period. Decomposing factors into three types of effects: effects and interim sales effects;
A computer program for causing a computer to execute the step of outputting information based on the result of the factorization,
Wherein the factor decomposing, if interest and dividend date is interest and a payout data transfers date the interest and dividend data includes as change data stocks were or purchased it holds belongs to the computation period, Calculate the initial holding effect and the interim purchase effect using the determination result and the interest / dividend data of the stock, and then subtract the initial holding effect and the interim purchase effect from the calculated overall rate of return. To calculate the effect of sale during the period on the memory ,
Computer program.

Images