Loss Financials Data Dictionary

The Loss Financials curated dataset is an opinionated view of transaction line items from Guidewire ClaimCenter. The dataset contains several derived transactional amounts that are generated by bucketing each line item into a type of recovery or expense, and then combining the reported amounts in these buckets to create new transactional amounts that represent different types of gross and net loss and expense. Additionally, the dataset contains several date-time constructs as well as policy, exposure, and claim attributes. These attributes and derived transactional amounts can be used by downstream applications to extract business intelligence for loss financial content.

ClaimCenter concepts

The transaction is the main financial entity in ClaimCenter. Every transaction contains one or more transaction line item objects that hold the all or part of the monetary amount. The amount of a transaction is the sum of all its transaction line items.

The transaction entity has the following subtypes:

The payment subtype represents money paid by the insurer to satisfy a claim. This amount includes a claim's liabilities, its associated Loss Adjustment Expenses (LAE), and other administrative expenses.
The recovery subtype represents money received by the insurer when a claim is settled. Salvage and subrogation are examples of recoveries.
The reserve subtype represents the estimated monetary amount that covers the insurer's future liabilities on a claim. Typically, payments decrease reserves.
The recovery reserve subtype represents the estimated money an insurer might recover when a claim is settled.

Grain of the dataset

The grain of the Loss Financials curated dataset is the transaction line item. At this lowest grain, payments and reserve changes are recorded for a variety of loss, expense and recovery categories.

The Bookingdate attribute of a Transaction entity is set when the transaction, which is immutable, is financially escalated in ClaimCenter, and the Life Cycle Status of the transaction is changed to Committed. From that point forward, all transaction line items that belong to the transaction inherit the Bookingdate.

The curation process for this dataset uses all transactions that have non-null values of the BookingDate field. For financial and statutory reporting purposes, all transaction line items are recognized to belong to the fiscal period to which the Bookingdate belongs. The dataset retains this field in its full timestamp format to facilitate proper sequencing of financial activity for a Claim or Exposure.

Based on the BookingDate, the dataset derives additional representations of the booking date such as Booking Year, Booking Quarter, and Booking Month.

How to use the dataset

The Loss Financials curated dataset contains raw and derived amounts from all transaction line items as they occur, without any aggregation across time. You can aggregate these line item metrics over several time periods — day, week, month, quarter, year, or Inception-to-Date (ITD):

You can group the data by BookingDate_Month, BookingDate_Quarter, or BookingDate_Year, to get Month-To-Date (MTD), Quarter-To-Date (QTD), and Year-To-Date (YTD) aggregated calendar year metrics.
You can use a WHERE clauses such as Booking Month <= YYYYMM or BookingDate <= MMDDYYYY to get ITD metrics as of that month or day.
You can sum up the transactions for a claim or exposure from the respective create time to any date that occurs before the current date to get the lifetime valuation of the claim or exposure. For open claims and exposures, you can get Paid ( Expenses, Losses and Recoveries) and Open Case Reserves (Expenses, Recoveries, and Losses). You can further decompose these amounts by various cost categories.

Buckets

The trans_alloc_code column in the dataset represents the buckets in the dataset. The curation process examines the values of subtype, costtype, and recoverycategory classifications of each line item and groups each line item into a bucket as follows:


`subtype`	`costtype`	`recoverycategory`	`trans_alloc_code`
`recovery`, `recoveryreserve`	`claimcost`	`credit_exp`, `credit_loss`	`indemnity_recovery`
		`deductible`	`deductible_recovery`
		`salvage`	`salvage_recovery`
		`subro`	`subro_recovery`
	`aoexpense`, `dcexpense`, `unspecified`	`credit_exp`, `credit_loss`	`expense_recovery`
	`aoexpense`	`salvage`	`salvage_expense_recovery`
	`unspecified`	`subro`	`subro_expense_recovery`
`payment`, `reserve`	`claimcost`	N/A	`medical` or `indemnity`, based on the `costcategory`.
	`aoexpense`		`lae_expense`
	`dcexpense`, `unspecified`		`ulae_expense`

Bucketed line item amounts

The reporting amount in a line item equals the transaction amount in the reporting currency. Based on the bucket in which the transaction line item falls, the curation process identifies the reporting amount in the line item as a bucketed line item. Bucketed line items are combined to determine various loss and expense amounts.

Terminology

dataset: Logical collection of related, discrete items of data. A single item can be associated with a single source table or with a primary source table and multiple secondary source tables.

Curated dataset: dataset that contains an enhanced view of the source data. Curated datasets are based on replicated datasets or other curated datasets. Source data is enhanced to add value, and includes joins, predicates (WHERE clauses), transformations, CDC, and other system columns.
Curation: Process of creating a curated dataset using ETL methods to combine, enrich, and transform source data. The source for curation is replicated datasets or a previously curated dataset.