Build the Portfolio
Enter properties, investments, insurance (and reinsurance) policies directly in the Engine. For larger books, download the CSV templates, complete them, and import the files in one step.
Product demo
Run your portfolio through Silfio's risk engine
This walkthrough covers a complete Silfio run. From uploading your portfolio, to running through the model, to finally interpreting model outputs.
engine.silfio.com/scenario-test
Iran stress case
A property and bond portfolio run against imposed country assumptions. Figures are illustrative.
Results ready
Annual conflict probability
80%
Buildings damaged
15%
GDP shock
-8%
Equity shock
-25%
Expected loss
$385K
TVaR 95%
$2.8M
VaR 99%
$4.1M
Hit frequency
18.4%
Engine walkthrough
Run using model best-estimates.
The Engine gives you the model's estimated view of losses on your portfolio. Using historical data, adjusted for the current geopolitical climate, the effects of war on your portfolio are simulated.
Run the Engine
Click Run Engine to send the portfolio to the model. The Engine validates every input, runs the simulations, and returns portfolio, property, investment, and conflict summaries.
Interpret the Model Outputs
Analyse the expected losses as well as the extreme cases (using VaR and TVaR metrics). These values are given on an asset level as well as on a portfolio level.
Scenario Test
Stress test scenarios with fixed assumptions.
Select countries, fix the assumptions that define your case, and let the model sample everything you leave open. Useful for stress testing, scenario analysis and for challenging the model's assumptions
Build the Portfolio
Enter properties, investments, insurance (and reinsurance) policies directly in Scenario Test. For larger books, download the CSV templates, complete them, and import the files in one step.
Set Scenario Parameters
Select one or more countries, then set conflict probability, building damage, GDP, inflation, equity, and bond-yield assumptions. Fields left blank are sampled by the model.
Run Scenario Test
Click Run Scenario Test to send the portfolio and scenario assumptions to the model. The scenario run returns the same output structure as a baseline Engine run.
Interpret the Model Outputs
Compare expected losses and tail losses against a baseline Engine run. Scenario outputs use the same VaR, TVaR, asset-level, and portfolio-level metrics.
Try this: Upload at least one property or investment from both Ukraine and Poland. Select Ukraine, fix annual conflict probability at 80% and physical damage to 30%, leave the economic fields blank for the model to generate. Now select Poland, leave the model annual conflict probability blank to take model estimate. Set physical damage to 7%, GDP shock to -5%, inflation shock to 3%, equity shock to -30% and leave bond yield change blank for the model to generate. Next run against the demo property and investment data. Inspect model results.
Interpreting outputs
Read the model from summary to detail.
Each run produces the same four summary files. Start with the portfolio view, then drill into the row-level files to view the individual breakdowns.
Portfolio summary
portfolio_summary.csv
Inventory summary
inventory_summary.csv
Assets summary
assets_summary.csv
Conflict summary
conflict_summary.csv
Loss Metrics
Expected loss
The average loss across all modelled years.
VaR and TVaR
VaR is a loss threshold. TVaR is the average loss inside the severe tail beyond that threshold.
API workflow
Automate the same workflow in your preferred language.
The API accepts the same exposure data as the browser workflow and returns the same portfolio, property, investment, and conflict summaries as CSV files inside a ZIP archive.
Use POST /run for baseline model estimates, or POST /scenario when you want to impose country-level assumptions.
Try it on your own portfolio.
Run a baseline Engine pass first, then move into Scenario Test when you want to impose assumptions and measure the difference.