Probable Maximum Loss (PML) Estimates
This dataset provides detailed Probable Maximum Loss (PML) estimates for various catastrophe exposures, including loss amounts at multiple return periods and confidence levels. It is designed to support reinsurance purchasing, risk management, and regulatory reporting by offering granular, traceable loss metrics across exposures and event types. The dataset is ideal for actuaries, risk managers, and insurance analysts seeking to quantify and manage catastrophic risk.
Sample rows
preview · 8 of 501 rows · all 11 columns| pml_idstring | catastrophe_typestring | confidence_level_percentfloat | reinsurance_layerstring | exposure_idstring | return_period_yearsinteger | pml_amountfloat | currencystring | calculation_datedate | model_versionstring | notesstring |
|---|---|---|---|---|---|---|---|---|---|---|
| PML-0001A | hurricane | 99.7 | excess of loss | US-FL-01 | 100 | 250000000 | USD | 2023-06-15 | HR2023.a | Peak hurricane risk for Florida, updated per 2023 model revision. |
| PML-0002B | earthquake | 99.9 | stop loss | JP-Kanto-02 | 1000 | 800000000 | JPY | 2022-11-30 | EQ2022.1 | Major Tokyo region earthquake scenario, high exposure portfolio. |
| PML-0003C | flood | 90 | quota share | IN-GJ-03 | 10 | 6500000 | INR | 2024-01-20 | FLD2024.c | Seasonal flooding in Gujarat, moderate portfolio exposure. |
| PML-0004D | wildfire | 95 | blank | AU-VIC-04 | 50 | 9000000 | AUD | 2021-08-10 | WF2021.3 | Typical wildfire risk, Victorian rural exposures. |
| PML-0005E | tornado | 90 | quota share | US-TX-05 | 20 | 3200000 | USD | 2023-04-01 | TOR2023.x | Moderate tornado scenario, North Texas region. |
| PML-0006F | flood | 95 | excess of loss | DE-BW-06 | 50 | 8200000 | EUR | 2022-07-12 | FLD2022.b | Flood risk assessment, Baden-Württemberg, updated after July 2022 event. |
| PML-0007G | flood | 0 | blank | BR-SP-07 | 1 | 0 | BRL | 2023-02-14 | blank | Edge case: no modeled loss for frequent São Paulo flooding. |
| PML-0008H | other | 10 | blank | ZA-GT-08 | 5 | 12000 | ZAR | 2024-03-18 | blank | Boundary scenario: low loss from recurring local event. |
| PML-0009I | hurricane | 99.5 | excess of loss | US-LA-09 | 100 | 275000000 | USD | 2021-09-25 | HR2021.b | Louisiana hurricane, high risk, excess of loss structure. |
| PML-0010J | flood | 99 | quota share | FR-IDF-10 | 100 | 14500000 | EUR | 2022-10-02 | FLD2022.2 | Seine region flooding, Paris metro, quota share treaty. |
| PML-0011K | hurricane | 99.99 | stop loss | MX-NLE-11 | 2500 | 1200000000 | MXN | 2023-05-22 | HR2023.c | Rare northern Mexico hurricane, ultra-long return period. |
| PML-0012L | earthquake | 100 | stop loss | NZ-AKL-12 | 5000 | 900000000 | NZD | 2021-02-19 | EQ2021.x | Auckland earthquake scenario, boundary risk case. |
| PML-0013M | earthquake | 99 | excess of loss | US-CA-13 | 100 | 340000000 | USD | 2024-02-07 | EQ2024.a | San Francisco Bay Area, peak earthquake risk. |
| PML-0014N | flood | 20 | blank | PK-PB-14 | 2 | 15000 | PKR | 2022-07-28 | blank | Low level flooding, Punjab, frequent event, not reinsured. |
| PML-0015O | wildfire | 80 | surplus share | RU-KR-15 | 30 | 7500000 | RUB | 2021-05-16 | WF2021.y | Wildfire risk, Karelia region, moderate exposure. |
| PML-0016P | tornado | 95 | blank | CA-ON-16 | 50 | 6200000 | CAD | 2023-07-03 | TOR2023.z | Ontario tornado, moderate risk, not reinsured. |
| PML-0017Q | flood | 99.5 | quota share | EU-PL-17 | 100 | 17500000 | EUR | 2022-03-06 | FLD2022.3 | Vistula region flooding, Polish exposures, quota share. |
| PML-0018R | flood | 0 | blank | SG-CE-18 | 1 | 0 | SGD | 2023-01-14 | blank | No modeled loss for frequent urban Singapore flooding. |
| PML-0019S | hurricane | 99.9 | excess of loss | US-AL-19 | 500 | 480000000 | USD | 2024-03-07 | HR2024.b | Alabama hurricane scenario, excess of loss program. |
| PML-0020T | earthquake | 99 | quota share | ES-MD-20 | 100 | 29500000 | EUR | 2023-12-24 | EQ2023.2 | Madrid region earthquake, moderate risk, quota share treaty. |
What the 501 rows show
from the 501-row sampleOther (catastrophe type) stands out: mean confidence_
- 91.0median confidence_
level_ percent - 16currencies
- 60median return_
period_ years - 4,100,000median pml_
amount
Median 91.0, from 0.0 to 100.0.
- string 7
- integer 1
- float 2
- date 1
Columns
11 columns in three groups| column | type | description | example |
|---|---|---|---|
| Text 7 columns | |||
pml_id | string | Unique identifier for each Probable Maximum Loss (PML) estimate recordunique | PML-0001A |
exposure_id | string | Identifier for the exposure or portfolio being analyzed (e.g., region, business unit, or risk group) | US-FL-01 |
catastrophe_type | string | Type of catastrophe event (e.g., earthquake, flood, hurricane)6 values | hurricane |
currency | string | Currency code for the PML amount (ISO 4217, e.g., USD, EUR, GBP) | USD |
model_version | string | Version identifier of the catastrophe model used for calculationoptional | HR2023.a |
reinsurance_layer | string | Reinsurance layer or program associated with the PML (e.g., 'excess of loss', 'quota share')5 layers · optional | excess of loss |
notes | string | Additional notes or comments regarding the PML estimateoptional | Peak hurricane risk for F… |
| Numbers 3 columns | |||
return_period_years | integer | Return period in years (e.g., 100 for 1-in-100-year event)1 or more | 100 |
confidence_level_percent | float | Confidence level for the PML estimate, expressed as a percentage (e.g., 99.5)0 to 100 | 99.7 |
pml_amount | float | Estimated probable maximum loss amount in the reporting currency0 or more | 250000000 |
| Dates and times 1 column | |||
calculation_date | date | Date when the PML calculation was performed | 2023-06-15 |
Use it for
An insurance dashboard
Confidence_
level_ percent by catastrophe_ type and a breakdown of reinsurance_ layer. Excel, Power BI or Tableau. Why do the 46 other rows have a mean confidence_
level_ percent of 3.1? A root-cause class exercise
Hand out the rows and one question. The answer is in the data, not in the brief.
- Pmls501PML-0001A99.7hurricanePML-0002B99.9earthqua…PML-0003C90flood
A software demo
Believable pmls with exposure_
id, catastrophe_ type and return_ period_ years to fill a screen in front of a buyer.
blueprint · probable-maximum-loss-pml-estimates
Behind this dataset
Same schema. As many rows as you need.
These 501 rows came out of a blueprint — 11 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.
- Return periods: 100-yr, 250-yr, 500-yr, 1000-yr events
- Confidence levels: 90%, 95%, 99%
- Peril-specific PMLs
- All-peril aggregate PML
- PML as percentage of total insured value (TIV)
- Occurrence PML vs aggregate annual PML
- PML used to determine reinsurance attachment points
- Regulatory PML reporting requirements
1 credit per row. Preview 20 rows free. Buy rows for the full run.
- Exports
- CSV, JSON, JSONL, Parquet, SQL, Excel, TSV, XML
- Licence
- yours to use, including commercially
- API slug
- probable-maximum-loss-pml-estimates