1. Upload Details

1.1 Data Requirements

Before the data is analyzed, the file must meet several basic requirements to ensure it can be read correctly.

1. The file must be an Excel file

   It must be one of the following: .xlsx, .xls, or .xlsb

2. The file must not be corrupted
3. The file cannot be empty
4. The file cannot exceed the 100 columns
5. Column A must contain data
6. Column headers must exist in Row 1

1.2 Data Cleaning

To guard against unintentional invalid dates entering the analysis, any date taking place in the year 1900 will be cleared to blank. This includes dates formatted 6/17/1900, Excel’s attempt at formatting 0 as a date: 1/0/1900. The year 1900 itself may still be included and will not be cleared.

2. Removing Contracts That Cannot Be Used

Before any analysis begins, certain contracts are removed entirely because they have illogical data

1. A Claim occurs before the Origination Date

   If any Claim Date is earlier than the Origination Date, the entire contract is excluded.

2. A Cancellation occurs before the Origination Date

   If a Cancellation Date appears before the Origination Date, the entire contract is excluded.

3. The Term is zero or negative

   Contracts must last at least one month to be meaningful. If the Term is not positive, the contract is removed.

4. Required contract elements are missing across all duplicates

   If a Contract ID is blank for any of the following:

   • Origination Date
   • Term
   • Premium

     then the entire contract is removed.

5. Records with no Contract ID are removed

If all Records are removed by these rules, the process stops and reports that no usable data remains.

---

3. How Dates After the Measurement Date Are Treated

Dates that occur after the Measurement Date are treated as events that have not yet happened:

1. Claims after the Measurement Date

   Any Claim Date later than the Measurement Date is treated as not existing

2. Cancellations after the Measurement Date

   Any Cancellation Date or Claim Date later than the Measurement Date is also treated as not having happened

---

4. How Duplicate Contracts Are Standardized and Used

A Contract ID may appear multiple times in the data. These are considered Duplicate Contracts.

4.1 Standardizing duplicates

To avoid contradictions between Duplicate Contracts, the system keeps the earliest available valid version of the following fields:

• Origination Date
• Term
• Premium
• Cancel Date

If later duplicates contain conflicting values, the earliest non-blank value is used for all records of that Contract ID.

If the first record with that Contract ID has a blank for that value, it can still be filled in if later duplicates have a non-blank value. I

4.2 How duplicates affect the analysis

• Duplicate Contracts are not counted twice.
• Claims in Duplicate Contracts are still counted as additional claims.

---

5. How Contracts Are Categorized

Each Contract ID is classified as of the Measurement Date into one of the following categories:

1. Cancelled

   The contract has a non-blank Cancellation Date that is after the Origination Date but before either the Maturity Date or the Measurement Date

2. Matured

   The Contract has reached its Maturity Date

   This is determined using 30 days * Term

3. Active

   The contract is still active as of the Measurement Date

---

6. How Premium Is Earned Over Time

Premiums are earned monthly according to the selected earning pattern. All methods calculate earned amounts through the Measurement Date.

6.1 Pro-Rata (Straight-Line)

• Premium is earned evenly each month across the entire Term.

6.2 Rule of 78s

• Premium is front-loaded.
• More of the Premium is earned in earlier months, with decreasing amounts later.

6.3 Reverse Rule of 78s

• Premium is back-loaded.
• Less Premium is earned in the early months and more near the end of the Term.

6.4 Custom Curve

• A Custom Curve can be uploaded specifying month-by-month incremental percentages.
• The sum of all increments must equal 100%.