Paid Loss Development Triangles

This dataset provides detailed cumulative paid loss development triangles segmented by accident year, development period, and line of business, enabling actuaries and analysts to track payment progression, estimate ultimate losses, and assess reserve adequacy. It includes reserve estimates, currency codes, and source information for robust actuarial modeling and regulatory reporting.

  • opened 5 times
  • last updated 2 Nov 2025
  • by GoMask
The brief that made it

Ultimate loss estimation for actuarial modeling

Sample rows

preview · 8 of 800 rows · all 10 columns
triangle_idstringcurrencystringcumulative_paid_lossfloatline_of_businessstringaccident_yearintegerdevelopment_periodintegeras_of_datedatedata_sourcestringreserve_estimatefloatnotesstring
TRI202401-0001USD120000.45Auto2017122018-01-31CoreClaims135000Initial 12 month reporting
TRI202401-0002USD91000.12Property2018122019-01-15LegacySystem95000blank
TRI202401-0003EUR200500Liability2019242021-01-10PartnerFeed210000Paid losses include subrogation
TRI202401-0004USD145000.89Auto2018242020-01-31CoreClaims165000blank
TRI202401-0005CAD184500.75Property2016362019-02-15LegacySystem190000Catastrophe event
TRI202401-0006GBP37000Specialty2017122018-01-10ExternalVendor42000blank
TRI202401-0007USD98000.5Auto2019122020-01-31CoreClaims105000blank
TRI202401-0008USD167500Liability2018242020-02-12PartnerFeed172000blank

What the 800 rows show

from the 800-row sample
  • 75,105median cumulative_paid_loss
  • 156data sources
  • 24median development_period
  • 84,000median reserve_estimate
currency800 rows by value
0245490489USD61%128EUR16%109GBP14%66CAD8%7AUD1%1JPY0%
cumulative_paid_loss800 rows, in bands of 200k
035571070175134430600k1.2Mcumulative_paid_loss →

Median 75,105, from 0.0 to 1,120,000.

line_of_business800 rows · 7 values
  1. Auto247
  2. Property243
  3. Liability230
  4. Workers Compensation36
  5. Marine23
  6. Specialty11
  7. General Accident10
10 columns by typefrom the column list below
  • string 5
  • integer 2
  • float 2
  • date 1

Columns

10 columns in three groups
blueprint · 10 columns
columntypedescriptionexample
Text 5 columns
triangle_idstringUnique identifier for each paid loss development triangle recorduniqueTRI202401-0001
line_of_businessstringInsurance line of business (e.g., Auto, Property, Liability)Auto
currencystringCurrency code for the paid loss amount (e.g., USD, EUR)6 currenciesUSD
data_sourcestringSource system or organization providing the dataoptionalCoreClaims
notesstringAdditional comments or context about the triangle recordoptionalCatastrophe event
Numbers 4 columns
accident_yearintegerYear in which the insured event (accident) occurred1,900 to 2,1002017
development_periodintegerNumber of months or years since the accident year (e.g., 12, 24, 36)0 or more12
cumulative_paid_lossfloatTotal amount paid for losses as of the development period for the accident year and line of business0 or more120000.45
reserve_estimatefloatEstimated ultimate reserve for the accident year and line of business as of the as_of_date0 or more · optional135000
Dates and times 1 column
as_of_datedateDate as of which the cumulative paid loss is reported2018-01-31

Use it for

  • median cumulat…75,105800 rowsmean cumulative paid …118.9kUSD105.5kEUR68.9kGBP95kCAD

    An insurance dashboard

    Cumulative_paid_loss by currency and a breakdown of line_of_business. Excel, Power BI or Tableau.

  • What pattern can you find in 800 rows?

    A class exercise

    Hand out the rows and one question. Everyone works from the same 800 rows.

  • A software demo

    Believable triangles with line_of_business, accident_year and development_period to fill a screen in front of a buyer.

Not quite right?

Make it yours.

Same 10 columns, your size and your rules. See 20 rows before you pay.

Preview 20 rows free

1,000 rows: $1.99 · 10,000: $12.99One-time. No subscription. All prices

This dataset800 rows10 columns
Yours1,000 rows10 columns

blueprint · paid-loss-development-triangles

Behind this dataset

Same schema. As many rows as you need.

These 800 rows came out of a blueprint — 10 columns with generation rules behind each one. Open it in Data Factory to retune a column, add your own, wire in foreign keys, and run it at the size you actually need.

Rules it was built with
  • Accident year as of loss occurrence date
  • Development period in months or years from accident year
  • Cumulative paid losses (not incremental)
  • Organized in triangle format: rows = accident years, columns = development periods
  • Longer-tail lines show development over 10+ years
  • Short-tail lines (property) develop quickly (12-24 months)
  • Age-to-age factors calculated from triangle
  • Tail factor for development beyond observed period
Rows
Open the blueprint in Data Factory

1 credit per row. New accounts start with 25 free credits.

Exports
CSV, JSON, JSONL, Parquet, SQL, Excel, TSV, XML
Licence
yours to use, including commercially
API slug
paid-loss-development-triangles

More in Insurance

Browse all
Also in InsuranceProbable Maximum Loss (PML) EstimatesMedical Claims Fraud ScenariosMedical Insurance Fraud DetectionAll 1,954 datasets

What should your data show?

Preview 20 rows free
No signup. No card.