Skip to end of metadata
Go to start of metadata

You are viewing an old version of this page. View the current version.

Compare with Current View Page History

« Previous Version 15 Next »

.

A floating weight blended benchmark calculates a blended index based on adjusting the blended weights of underlying segments or sectors to reflect index performance. You can define static weights by entering initial weights for the custom benchmark, or active weights by identifying a reference entity to use as the source of weights for the custom benchmark. The custom benchmark (target benchmark) return equals the weighted sum of the vendor index (source index) returns. The weights are calculated by PACE, in between reset dates, using weights that reflect the relative change in source index levels over the period.

For example, 50 percent LEHMAN + 50 percent S&P500, where the weights are reset to 50 percent quarterly and allowed to float with the market in between quarters. Reset dates are configured and set using a date rule and a business calendar.

Floating type custom benchmarks include all functionality available with Blended type benchmarks. Float type custom benchmarks additionally offer you the option to define a reset frequency for rebalancing weights and to apply weights from another entity. If you are creating a custom benchmark that requires current blended benchmark functionality, you can set up the custom benchmark as a Floating type benchmark to provide greater flexibility for the benchmark in the future.

Calculate Weights and Returns

You can configure a blended benchmark so that PACE determines the weights to use, where the weights are calculated based on the component index returns between reset dates. This type of benchmark is called floating weight blended benchmark. An example of how PACE calculates the weights and returns for these types of benchmarks follows.

Suppose blended benchmark weights for December month end are used to calculate the January return for the custom blended benchmark. See the following figure.

On this page:


December Benchmark Weights to Calculate January Return

Assume that the weights for February month end are to float based on the market changes.

First, calculate the February month-end returns using the January month-end weights. The January month-end weights are calculated by referencing the December month-end weights and the January returns. See the following figure.

January Month-End Weights Used to Calculate February Return

The January month-end weights are calculated by growing the December weight by the January return, and then re ratioing the resulting weights to 100 percent.

The weights used in the calculation are those that are stored in the ABAL column on the PERF_SEC_RETURNS table. The returns used are those that are stored in the PSR_USER_FLOAT8 column (or the TOT_RETURN column, for clients who installed Eagle Performance prior to V11.0).

Each time the system builds the floating weight custom benchmark, it inserts a new custom index attributes definition record in the Custom Index Attributes table for that effective date.

The PACE Floating Weight Calculator uses the following methodology to populate the RULES…CUSTOM_INDEX_ATTRIBUTES table:

  • It finds the constituent benchmarks of the Custom Benchmark and Date for Return from the most recent custom index attributes definition record in the Custom Index Attributes table.

The daily or monthly option is chosen during the submission of the entity build. The date for the return as mentioned here is the last date for the specific entity in the custom index attributes table. There is no daily or monthly functionality at this point in the process.

  • It finds the return for the benchmarks for the date from the previous step.

The return is the PERFORM..PERF_SEC_RETURNS.PSR_USER_FLOAT8. (This field is a default but you can alter it. It is set up using an internal field attribute 325 called i Blended BNCH Wght Col. This is created using the scripts — you do not need to create this field attribute. The floating return can be changed from the default of PSR_USER_FLOAT8 to another return field by running the following script: Update rules..field_attributes set field_name = mod_dietz_return* where field_attribute_id = 325.
In this example, mod_dietz_return is the new return field to be used.)

  • It finds the current weights from the most recent previous custom index attributes definition (to the entity build date) in the Custom Index Attributes table.

Next, Eagle populates the Custom Index Attributes table for this effective date with the new weights that it calculated according to the math described below. When it creates a new record in the Custom Index Attributes table for the custom index attributes definition, it copies the previous record's data, and applies the newly calculated weights. It populates other fields in the record just as they appeared in the previous custom index attributes definition from the Custom Index Attributes table, with the exception of the update user, update time, and instance fields.

Finally, after the system populates the Custom Index Attributes table, the entity build procedure begins. These two steps cannot be separated.

Rebalance Floating Weights Using Reset Dates

Behind the scenes, the system uses the following two columns in the Custom Index Attributes table for resetting the dates:

  • RESET_DATE_ID holds the date rule ID that is associated with the floating benchmark.
  • RESET_FLAG is used by the Eagle system to indicate a reset value.

On user-designated reset dates (which you must identify using a Business Calendar) the system resets weights that are floating. On each reset date, the Eagle system looks for the most recently dated custom index attributes definition with a RESET_FLAG that is set to Y and takes the weights from this definition without floating them. A custom index attributes definition with a reset flag of Y indicates it is a reset value, and the system can apply its weights for reset purposes on the next reset date.

When does a custom index attributes definition have a RESET_FLAG set to Y? The system sets the custom index attributes definition's RESET_FLAG to Y when:

  • You add a new Custom Index Attributes Definition. When you first create a floating benchmark, you create a new custom index attributes definition. When you enter information in the Custom Index Attributes tab or dialog box in effect for a specific date, the system creates a new custom index attributes definition and sets the RESET_FLAG to Y. (For details, see Create a Floating Weight Blended Benchmark.
  • You change a Custom Index Attributes Definition. When you modify a floating benchmark and you edit a custom index attributes definition that already exists for a specified date, the system sets the custom index attributes definition's RESET_FLAG to Y if you change any of the following in the Custom Index Attributes tab or dialog box: the date rule, the weights, the source entity, or the source dictionary.

You cannot change the target dictionary and the type of custom benchmark in the Custom Index Attributes tab after you create the entity. These items are dimmed in the Custom Index Attributes tab.

Floating Weight Blended Benchmark Example

The following example shows how the floating weight blended benchmark is calculated:

  1. Calculate the drift numbers.
    Index 1  (index rtn/100) * current weight = drift #1 (15/100) * 50 = 7.5
    Index 2  (index 2 rtn/100) * current weight = drift #2 (-10/100) * 50 = -5

  2. Calculate the new weights.
    Previous weight (current weight from previous step) + Drift #1 = New weight Index 1 50 + 7.5 = 57.5
    Previous weight (current weight from previous step) + Drift #2 = New weight Index 2 50 + (-5) = 45
    New weight Index 1 + New weight Index 2 = Total new weight 57.5 + 45 = 102.5

  3. Return the numbers back to a 100 percent ratio.
    Weights must equal 100 for the custom index builder process.
    New weight Index 1/Total New weight * 100
    57.5/102.5 = x/100 x = 56.1
    New weight Index 2/Total New weight * 100
    45/102.5 = x/100 x= 43.9

3-Month Floating Weight Blended Benchmark Example

Index returns for this example are listed in the following table. Calculations for this data are described below.


January

February

Benchmark 1

-2%

9%

Benchmark 2

-6%

4%

Benchmark 3

1%

3%

Month 1 Floating Weight Calculation Nonreset (February) Example

In the example defined in the following table, weights for 2/28/2002 are calculated using weights and returns from 1/31/2002.


Effective Date

Entity Detail ID

Re-basing Calc

Weight

Original Weights

1/31/2002 0:00

Benchmark1


65


1/31/2002 0:00

Benchmark2


25


1/31/2002 0:00

Benchmark3


10





100


2/28/2002 0:00

Benchmark1

63.7000

65.467626


2/28/2002 0:00

Benchmark2

23.5000

24.152107


2/28/2002 0:00

Benchmark3

10.1000

10.380267




97.3000

100

Calculation details for this example are shown in the following table.

Benchmark

Calculation

Step 1 – Drift numbers


Benchmark 1

(0.02) * 65 = (-1.3)

Benchmark 2

(0.06) * 25 = (-1.5)

Benchmark 3

0.01 * 10 = 0.10

Step 2 – New weights


Benchmark 1

(-1.3) + 65 = 63.7

Benchmark 2

(-1.5) + 25 = 23.5

Benchmark 3

0.10 + 10 = 10.1

Total

63.7 + 23.5 + 10.1 = 97.3

Step 3 – Ratio back to 100%


Benchmark 1

63.7/ 97.3 * 100 = 65.467626

Benchmark 2

23.5/ 97.3 * 100 = 24.152107

Benchmark 3

10.1/ 97.3 * 100 = 10.380267

Month 2 Floating Weight Calculation Nonreset (March) Example

In the example defined in the following table, the date of 3/31/2002 is calculated using the weights and the return for 2/28/2002.


Effective Date

Entity Detail ID

Re-basing Calc

Weight

Original Weights

1/31/2002 0:00

Benchmark1


65


1/31/2002 0:00

Benchmark2


25


1/31/2002 0:00

Benchmark3


10





100


2/28/2002 0:00

Benchmark1

63.7000

65.467626


2/28/2002 0:00

Benchmark2

23.5000

24.152107


2/28/2002 0:00

Benchmark3

10.1000

10.380267




97.3000

100


3/31/2002 0:00

Benchmark1

71.359712

66.585792


3/31/2002 0:00

Benchmark2

25.118191

23.437800


3/31/2002 0:00

Benchmark3

10.691675

9.976409




107.169578

100

Calculation details for this example are shown in the following table.

Benchmark

Calculation

Step 1 – Drift numbers


Benchmark 1

0.09 * 65.467626 = 5.892086

Benchmark 2

0.04 * 24.152107 = 0.966084

Benchmark 3

0.03 * 10.380267 = 0.311408

Step 2 – New weights


Benchmark 1

65.467626 + 5.892086 = 71.359712

Benchmark 2

24.152107 + 0.966084 = 25.118191

Benchmark 3

10.380267 + 0.311408 = 10.691675

Total

71.359712 + 25.118191 + 10.691675 = 107.169578

Step 3 – Ratio back to 100%


Benchmark 1

71.359712 / 107.169578 * 100 = 66.585792

Benchmark 2

25.118191 / 107.169578 * 100 = 23.437800

Benchmark 3

10.691675 / 107.169578 * 100 = 9.976409

Month 3 Floating Weight Calculation Reset (April) Example

Given that 4/30/2002 is a reset date as set up in the business calendar, then the weights for 4/30 are equal to the weights on the last reset date of 1/31/2002. The example is described in the following table.


Effective Date

Entity Detail ID

Re-basing Calc

Weight

Original Weights

4/30/2002 0:00

Benchmark1


65


4/30/2002 0:00

Benchmark2


25


4/30/2002 0:00

Benchmark3


10





100


Configure a Floating Weight Blended Benchmark

To configure a floating weight blended benchmark:

  1. Set up the Business Calendar to identify reset dates for the benchmark.
  2. Set up the date rule.
  3. Create the new entity for the floating weight blended benchmark.

Set Up the Business Calendar

Weights float until you specify a reset date. A reset date is a date on which the system no longer floats from previous values but resets to the original or last updated weights for the benchmark in the Custom Index Attributes dialog box. That is, on the reset day, the weights reset to those weights associated with the most recent date for the custom benchmark definition where the RESET_FLAG = Y in the CUSTOM_INDEX_ATTRIBUTES table for that specific entity. For more information about the RESET_FLAG, see Rebalance Floating Weights Using Reset Dates.

To specify a reset date, you use the business calendar. Eagle recommends that you keep the source and business calendar used for resets separate from other business calendars so that the reset functionality is not dependent on another calendar's dates.

To set up reset dates for a floating weight blended benchmark:

  1. From any Eagle window, click the Eagle Navigator button to access the Eagle Navigator.
  2. Enter Sources in the Start Search text box.
  3. Click the Sources (Reporting Center) link to access the Sources workspace.
    You see the Sources window.
  4. Click Add.
    The Add Source Details dialog box appears.
  5. Add a new interface (source) with a meaningful name such as Source for Floating Weight Reset Dates. Under Available Feed Types, select Performance. See the following figure.
    Editing Source Details Dialog Box
  6. Update the Business Calendar for the source from any Eagle window, click the Eagle Navigator button to access the Eagle Navigator. Enter Business Calendar in the Start Search text box.
  7. Click the Business Calendar (System Management Center) link to access the Business Calendar workspace.
  8. Choose the source created for the floating weight blended benchmarks and select the check boxes, under the Daily column, for the reset dates. Make sure to use the Daily column — the program only uses this column. For example, if you reset on a quarterly basis, then only select four dates with a check mark.
  9. Click Save before exiting.

Set Up the Date Rule

You can set up the date rule for a floating weight blended benchmark. is located in PACE Reporting under the General Reporting option.

To set up the date rule for a floating weight blended benchmark:

  1. From any Eagle window, click the Eagle Navigator button to access the Eagle Navigator.
  2. Enter Date Rules in the Start Search text box.
  3. Click the Date Rules (Reporting Center) link to access the Date Rules workspace.
  4. Select New.

    Set Up the Date Rule example
  5. Create a new date rule like the one shown in the previous figure as follows.
    In step 1, the Name field, enter a meaningful name, such as Floating Reset Rule.
    In steps 3 and 4, choose to calculate using the business calendar.
    In step 6, use the business calendar source that you created in the previous procedure. Note that the business calendar appears under its long name. The business calendar is unavailable unless you select the Using Business Calendar check box.

    Return Column

PACE system variable, i Blended BNCH Wght Col (PACE system id- 325), points to the column that PACE uses as the return column in the source entity.

Reference Entity: Eagle Balance Fund





Investment Type

Security Type

Industry Type

Weight in Ref Entity on 3/31/2007 (%)

Custom Index Attribute

EQ



72



Common


40

Total --> Common S&P 500



Finance

20




Technology

10




Health Care

10



Preferred


22

Total --> Preferred Russell 2000



Finance

8




Technology

2




Health Care

12



144 A


10




Health Care

10

Total --> 144A --> Health care ML HealthCare Index

FI



28



CORP


12

Total -> Corp L AGG



Finance

3




Manufacturing

3




Health Care

3




Energy

3



GOVT


8

Total -> Gov L AGG



Sovereign

3




Agency

5



MBS


2

Total -> MBS L AGG



MBS

2



MUNI


4

Total -> L MUNI Index



Federal

1




State

1




Town

2


If you clear the Do not build Target Dictionary below assigned level check box, the node structure below the selected node has to match the target dictionary for the entity build to run successfully. See Build the Target Dictionary at Assigned Level and Above.


Partially Filled Dictionary

Optimally, you would assign a benchmark for each node of the reference entity. But this is not always the case. It is conceivable that benchmarks will not be assigned to some of the Reference entity nodes. Consider the example in the following table.

Investment Type

Security Type

Industry Type

Weight in Ref Entity on 3/31/2007(%)

Custom Index Attribute

EQ



72



Common


40

Total --> Common S&P 500



Finance

20




Technology

10




Health Care

10



Preferred


22




Finance

8




Technology

2




Health Care

12



144 A


10




Health Care

10

Total --> 144A --> Health Care ML Health Care Index

FI



28



CORP


12

Total -> Corp L AGG



Finance

3




Manufacturing

3




Health Care

3




Energy

3



GOVT


8

Total -> Gov L AGG



Sovereign

3




Agency

5



MBS


2




MBS

2



MUNI


4




Federal

1




State

1




Town

2


The custom benchmarks from the previous example only use the weights and nodes from the reference entity:

  • Total -->  Common S&P 500 – 40%
  • Total --> 144A --> Health Care ML Health Care Index – 10%
  • Total  --> Corp L AGG – 12%
  • Total --> Gov L AGG – 8%

Change Weights on an Ad Hoc Basis after Initial Setup

In most cases, you need to manually enter weights only when you first set up the floating weight benchmark. After that, the system calculates the floating weights for you, and the system resets weights automatically at the appropriate times. However, if you need to manually change a weight on an ad hoc basis at a later date, you can do so.

Because the system is designed to overwrite weights each time a day is calculated, enter any ad hoc weight change on a reset date (as specified in the business calendar).When the system rebalances weights on a reset date, it uses the most recently dated custom index attributes definition with a RESET_FLAG that is set to Y and takes the weights from this date.

To manually change a weight on an ad hoc basis:

  1. Edit the floating weight benchmark and enter the new weight.
    You can modify an existing custom index attributes definition to change the weight in effect on a custom index attributes definition you already created or can add a new custom index attributes definition to enter a weight in effect on a new date not already associated with a custom index attributes definition.
  2. Ensure that the business calendar used for resets includes a reset date for the weight change.
    If that business calendar does not already include the reset date for the date when the weight change goes into effect, add a reset date for the ad hoc weight change to the business calendar.
  3. Recalculate performance to reflect the weight change.
    If the reset date associated with the weight change is backdated, you must recalculate performance for that backdated reset date and going forward through to the current day. The weights from the ad hoc date are marked with the RESET_FLAG = Y and the system can apply them for reset purposes on the next reset date. 

  • No labels