Where a counterparty uses an initial margin model, that model may be developed by any of, or both, counterparties or by a third-party agent.

Where a counterparty uses an initial margin model developed by a third-party agent, the counterparty shall remain responsible for ensuring that that model complies with the requirements referred to in this Section.

Initial margin models shall be developed in a way that captures all the significant risks arising from entering into the non-centrally cleared OTC derivative contracts included in the netting set, including the nature, scale, and complexity of those risks and shall meet the following requirements:

1. (a)

   the model incorporates risk factors corresponding to the individual currencies in which those contracts in the netting set are denominated;

2. (b)

   the model incorporates interest rate risk factors corresponding to the individual currencies in which those contracts are denominated;

3. (c)

   the yield curve is divided into a minimum of six maturity buckets for exposures to interest-rate risk in the major currencies and markets;

4. (d)

   the model captures the risk of movements between different yield curves and between different maturity buckets;

5. (e)

   the model incorporates separate risk factors at least for each equity, equity index, commodity or commodity index which is significant for those contracts;

6. (f)

   the model captures the risk arising from less liquid positions and positions with limited price transparency within realistic market scenarios;

7. (g)

   the model captures the risk, otherwise not captured by other features of the model, arising from derivative contracts where the underlying asset class is credit;

8. (h)

   the model captures the risk of movements between similar, but not identical, underlying risk factors and the exposure to changes in values arising from maturity mismatches;

9. (i)

   the model captures main non-linear dependencies;

10. (j)

    the model incorporates methodologies used for back-testing which include statistical tests of the model's performance;

11. (k)

    the model determines which events trigger a model change, calibration or other remedial action.

The risk management procedures referred to in Article 2(1) shall ensure that the performance of the model is monitored on a continuous basis including by back-testing the model at least every 3 months.

For the purposes of the first subparagraph, back testing shall include a comparison between the values produced by the model and the realised market values of the non-centrally cleared OTC derivative contracts in the netting set.

The risk management procedures referred to in Article 2(1) shall outline the methodologies used for undertaking back-testing, including statistical tests of performance.

The risk management procedures referred to in Article 2(1) shall describe what results of the back-testing would lead to a model change, recalibration or other remediation action.

The risk management procedures referred to in Article 2(1) shall ensure that counterparties retain records of the results of the back-testing referred to in paragraph 3 of this Article.

Counterparties shall provide all the information necessary to explain the calculation of a given value of the initial margin model to the other counterparty in a way that a knowledgeable third party would be able to verify that calculation.

The initial margin model shall reflect parameter uncertainty, correlation, basis risk and data quality in a prudent manner.

The assumed variations in the value of the non-centrally cleared OTC derivative contracts within the netting set for the calculation of initial margins using an initial margin model shall be based on a one-tailed 99 percent confidence interval over a MPOR of at least 10 days.

The MPOR for the calculation of initial margins using an initial margin model referred to in paragraph 1 shall include:

1. (a)

   the period that may elapse from the last margin exchange of variation margin to the default of the counterparty;

2. (b)

   the estimated period needed to replace each of the non-centrally cleared OTC derivative contracts within the netting set or hedge the risks arising from them, taking into account the level of liquidity of the market where those types of contracts are traded, the total volume of the non-centrally cleared OTC derivative contracts in that market and the number of participants in that market.

Parameters used in initial margin models shall be calibrated, at least annually, based on historical data from a time period with a minimum duration of 3 years and a maximum duration of 5 years.

The data used for calibrating the parameters of initial margin models shall include the most recent continuous period from the date on which the calibration referred to in paragraph 1 is performed and at least 25 % of those data shall be representative of a period of significant financial stress ("stressed data").

Where stressed data referred to in paragraph 2 does not constitute at least 25 % of the data used in the initial margin model, the least recent data of the historical data referred to in paragraph 1 shall be replaced by data from a period of significant financial stress, until the overall proportion of stressed data is at least 25 % of the overall data used in the initial margin model.

The period of significant financial stress used for calibration of the parameters shall be identified and applied separately at least for each of the asset classes referred to in Article 17(2).

The parameters shall be calibrated using equally weighted data.

The parameters may be calibrated for shorter periods than the MPOR determined in accordance with Article 15. Where shorter periods are used, the parameters shall be adjusted to that MPOR by an appropriate methodology.

Counterparties shall have written policies setting out the circumstances triggering a more frequent calibration.

Counterparties shall establish procedures for adjusting the value of the margins to be exchanged in response to a change in the parameters due to a change in market conditions. Those procedures shall provide for counterparties to be able to exchange the additional initial margin resulting from that change of the parameters over a period that ranges between 1 and 30 business days.

Counterparties shall establish procedures regarding the quality of the data used in the model in accordance with paragraph 1, including the selection of appropriate data providers and the cleaning and interpolation of that data.

Proxies for the data used in initial margin models shall be used only where both of the following conditions are met:

1. (a)

   available data is insufficient or is not reflective of the true volatility of an OTC derivative contract or portfolio of OTC derivative contracts within the netting set;

2. (b)

   the proxies lead to a conservative level of margins.

Initial margin models shall only include non-centrally cleared OTC derivative contracts within the same netting set. Initial margin models may provide for diversification, hedging and risk offsets arising from the risks of the contracts within the same netting set, provided that the diversification, hedging or risk offset is only carried out within the same underlying asset class as referred to in paragraph 2.

For the purposes of paragraph 1, diversification, hedging and risk offsets may only be carried out within the following underlying asset classes:

1. (a)

   interest rates, currency and inflation;

2. (b)

   equity;

3. (c)

   credit;

4. (d)

   commodities and gold;

5. (e)

   other.

Counterparties shall establish an internal governance process to assess the appropriateness of the initial margin model on a continuous basis, including all of the following:

1. (a)

   an initial validation of the model by suitably qualified persons who are independent from the persons developing the model;

2. (b)

   a follow up validation whenever a significant change is made to the initial margin model and at least annually;

3. (c)

   a regular audit process to assess the following:

   1. (i)

      the integrity and reliability of the data sources;

   2. (ii)

      the management information system used to run the model;

   3. (iii)

      the accuracy and completeness of data used;

   4. (iv)

      the accuracy and appropriateness of volatility and correlation assumptions.

The documentation of the risk management procedures referred to in point (b) of Article 2(2) relating to the initial margin model shall meet all of the following conditions:

1. (a)

   it shall allow a knowledgeable third party to understand the design and operational detail of the initial margin model;

2. (b)

   it shall contain the key assumptions and the limitations of the initial margin model;

3. (c)

   it shall define the circumstances under which the assumptions of the initial margin model are no longer valid.

Counterparties shall document all changes to the initial margin model. That documentation shall also detail the results of the validations, referred to in paragraph 1, carried out after those changes.