CCPs Revise Margin Models | Esma

According to a Risk Magazine article (subscription required), some European CCPs are revising their margin models in order to comply with Esma regulations regarding the effects of procyclicality on margin requirements.  
January 30, 2014 - Editor
Category: Clearing

According to a Risk Magazine article (subscription required), some European CCPs are revising their margin models in order to comply with Esma regulations regarding the effects of procyclicality on margin requirements.

 

According to a Risk Magazine article (subscription required), some European CCPs are revising their margin models in order to comply with Esma regulations regarding the effects of procyclicality on margin requirements.

Though the impact of margin volatility is experienced by many products, the effects are most pronounced for interest rate futures given their relatively large notional values.

ESMA has outlined and documented three methods (PDF) by which European CCPs can adapt their margin models to minimize procyclicality effects. The methods are as follows: 

  1. Add on a buffer of 25% to the minimum calculated margin requirement.
    • This has been implemented by LCH.Clearnet for the Nasdaq NLX platform. LCH have stated that this method of reducing procyclicality has the added benefit of providing a suitable stress condition for products where it's difficult, or impossible, to obtain historical data for their VaR-type simulation model.
  2. Assign a weight of at least 25% to the stress observations in the calculated lookback period. 
  3. Ensure that the margins are no lower than those calculated considering a 10-year lookback period.
    • This approach is favoured by Eurex. They have stated that it allows for margin models to be flexible whilst establishing the margin calculated using a 10-year lookback period as a floor.

For the most part, futures margin models based on historical VaR (H-VaR) are most sensitive to procyclicality effects. As explained in an earlier post comparing VaR and SPAN, if the oldest scenarios in an H-VaR model took place during a period of significant financial stress, then the daily difference between VaR calculations (and hence margin requirements) will be volatile as these scenarios "roll off" the model.

As the Risk article explains, models based on SPAN are generally less sensitive to procyclicality concerns, as CCPs have flexibility in changing the SPAN parameters to subjectively take these concerns into account. 

-Ben L.


Popular
Most Viewed

Image