How filtered historical simulation informs margin.

Filtered historical simulation, or FHS, uses historical risk-factor moves adjusted for volatility to build scenarios for a portfolio. It is one part of several margin methodologies. Understanding that role helps explain why a changing market or a different portfolio can change the result.

Discuss your margin workflow

From historical moves to portfolio risk

  1. Step 1

    Filter historical moves

    A model starts from historical risk-factor returns and adjusts them for volatility. This allows observed moves from different market conditions to inform scenarios under the model’s current volatility treatment.

  2. Step 2

    Revalue and aggregate

    Scenario risk factors produce scenario instrument values and profit or loss. Position quantities turn these into portfolio outcomes, with aggregation constrained by the model’s permitted groups and offsets.

  3. Step 3

    Measure tail risk

    The model applies its chosen risk measure to the scenario loss distribution. The resulting market-risk component may then be combined with stress measures and other adjustments under that specific methodology.

Filtering changes how history is used

An unfiltered historical sample carries the volatility of the periods it contains. FHS adjusts historical returns so the scenarios reflect the model’s volatility assumptions. The lookback, filtering rule, holding period and risk measure all affect the outcome, so “FHS” alone does not specify a complete margin calculation.

A market-risk estimate is not always total initial margin

ICE describes IRM 2 as a portfolio model based on FHS with additional stress and liquidity features. ECC’s published Prisma for Commodities methodology combines filtered-history and stress-period measures with model adjustments. The details differ: a formula or parameter choice from one model should not be assumed to apply to another.

  • Confirm which clearing house, model version and products apply.
  • Distinguish the market-risk component from the complete reported margin.
  • Check liquidity, concentration and other adjustments in the relevant methodology.

Keep methodology and product availability separate

Cumulus9 implements supported named margin methodologies, including Eurex PRISMA and ICE IRM 2 workflows. Their calculations use the applicable exchange inputs. This guide does not offer a universal FHS calculator, reproduce an exchange’s entire calibration process or establish support for every product using FHS.

Questions about FHS and margin

Is FHS another name for Value-at-Risk?
No. FHS describes how scenarios are constructed from historical data. Value-at-Risk is a risk measure that can be applied to the resulting loss distribution. A specific model defines how these pieces fit together.
Will a diversified portfolio always require less margin?
No. Recognised offsets depend on the model, product grouping and account treatment. Stress measures, concentration or other adjustments can limit an apparent benefit. Calculate the actual portfolio on a consistent basis.
Why can margin change when positions stay the same?
Prices, volatility, scenario inputs and other model parameters can change. Compare the calculation dates and parameter sets before attributing the difference to the portfolio itself.

References

Contact · request a demo

Run your book through it.

Bring us a portfolio and the scenarios that matter. We run it live, on your numbers — no commitment, and no data retention beyond the demo.

Postal address
30 St. Mary Axe, London, EC3A 8BF
Phone number
+44 (0) 20 7205 4090
Email
info@cumulus9.com

Replies within one business day