Module 3

Historical Performance

RiskSpan Edge Training Program

Learning Objectives

By the end of this module, you will be able to:

Understand what Historical Performance analysis provides
Select appropriate datasets for analysis
Configure output options and filters
Use bucketing for comparative analysis
Run and export reports

What is Historical Performance?

Definition

Historical Performance refers to the past behavior and results of financial instruments over a specific period.

                            For Mortgage Loans, This Means:
                            How loans have performed in terms of prepayments (CPR)
How loans have performed in terms of defaults (CDR)
Delinquency patterns over time
Credit quality trends

                        

Why It Matters

Historical data helps predict potential future risks and returns for similar investments.

Module Overview

Historical Performance Page Layout

Section	Contents
Left Pane	Module list (navigation)
Top Section	Dataset selection, Output options, Action buttons (Run Report, Add Series, Save CSV)
Middle Section	Performance Period settings, Quick Filters
Bottom Section	Advanced Filters/Bucketing

Selecting a Dataset

Pool Level Datasets

Code	Agency
FNMA	Fannie Mae pools
FHLMC	Freddie Mac pools
GNMA1, GNMA2	Ginnie Mae pools

Loan Level Datasets

Code	Type
FNLOAN	Fannie Mae loans
FHLOAN	Freddie Mac loans
GNLOAN, G2LOAN	Ginnie Mae loans
FNARMLOAN, FHARMLOAN	ARM loans

Multiple Datasets

Check "Bucket Data Set" to analyze multiple datasets simultaneously with results in separate columns.

Output Options

Choose Your Output Level

Option	Description	Best For
Quick Output	Limited essential info	Fast overview
Standard Output	Default data set	Regular analysis
Expanded Output	Detailed percentages	Deep dive research
Average CPR	3/9/12 month averages	Trend analysis

Tip: Click the radio button for your desired output type before running the report.

Performance Period Settings

Define Your Time Frame

                            Date Range
                            Start Year + Month
End Year + Month
OR specific Factor Date

                        

                            Step Size (Optional)
                            Expressed in months
Controls granularity of output

                        

                            Example
                            
                                        Start:
                                        January 2020
                                    
                                        End:
                                        December 2023
                                    
                                        Step:
                                        3 (quarterly buckets)

Quick Filters

Rapidly Narrow Results

Filter	Purpose
Pool Term	Filter by loan term (15yr, 30yr, etc.)
Pool Type	Filter by security type
Curr Bal >	Minimum current balance threshold
Num of Loans >	Minimum loan count

                    Bucketing Options
                    Bucket Loan Term: Show each term on separate row
Bucket Loan Type: Show types in distinct columns

                

Numeric Bucketing

Filtering by Number Ranges

For numeric fields (LTV, FICO, WAC, etc.):

                            Parameters
                            Min: Lower bound of range
Max: Upper bound of range
Step: Divides range into partitions

                        

Example: LTV Bucketing

Min: 60, Max: 80, Step: 10

Results in:

60 < LTV ≤ 70
70 < LTV ≤ 80

                            Important: "Left-Open, Right-Closed" Logic
                            Lower bound is exclusive (except lowest endpoint)
Upper bound is inclusive

                        

Categoric Bucketing

Filtering by Categories

For text/category fields (OccupancyType, Channel, etc.):

                            How It Works
                            Select the field to bucket
Choose specific values from dropdown
Results filter to selected categories

                        

Example

Field: OccupancyType

Value: "INVESTOR"

This filters to only investor-occupied properties.

Screenshot: Categoric Bucket Configuration

Bucket, Complement & All Options

Three Checkboxes for Each Filter

Option	What It Shows
Bucket	Separate row for each selected value
Complement	Everything EXCEPT selected values
All	Complete dataset (filter off)

Example with 3 Values Selected

If you select "WELLS FARGO", "CHASE", "BOFA" as servicers:

Bucket checked: 3 rows (one per servicer)
Bucket + Complement: 4 rows (3 servicers + "everything else")
Bucket + Complement + All: 5 rows (3 servicers + complement + total)

Running Reports

Step-by-Step Process

                            Before Running
                            Select Dataset from dropdown
Choose Output Type (Quick/Standard/Expanded/Avg CPR)
Set Performance Period (date range + optional step)
Apply Filters as needed
Click "Run Report"

                        

                            After Running
                            View results in grid below
Add Series: Compare with another cohort
Save CSV: Export data for offline analysis

                        

Interpreting Results

Key Output Columns

Column	Meaning
Factor Date	As-of date for the data
Balance	Total unpaid principal balance
Loan Count	Number of loans in cohort
CPR	Conditional Prepayment Rate
CDR	Conditional Default Rate
WAC	Weighted Average Coupon
WALA	Weighted Average Loan Age

                    Reading CPR/CDR
                    CPR 8.5 means 8.5% annual prepayment rate
CDR 0.5 means 0.5% annual default rate

                

Module 3 Summary

Key Takeaways

Historical Performance analyzes past CPR and CDR trends
Datasets include pool-level and loan-level for all agencies
Output options: Quick, Standard, Expanded, Average CPR
Numeric bucketing uses "left-open, right-closed" logic
Bucket/Complement/All provide flexible comparison views
Run Report executes the query; Save CSV exports results

Next Module

Module 4: Portfolio Module - Manage and analyze securities portfolios

Start:	January 2020
End:	December 2023
Step:	3 (quarterly buckets)