Margin settlement risk and its effect on CVA

Sponsored feature: CompatibL

Margin settlement risk and its effect on CVA

Until recently, the discourse on XVA had focused on new valuation adjustments: funding valuation adjustment (FVA), margin valuation adjustment (MVA) and capital valuation adjustment (KVA). Plain old credit valuation adjustment (CVA) seemed boring and devoid of surprises by comparison. This has changed with a series of recent publications from quants at Bank of America Merrill Lynch, the US Federal Reserve Board and CompatibL [1, 2, 3] and two influential editorials authored by Nazneen Sherif in Risk magazine. [4, 5] These publications highlight the importance of margin settlement risk and its major influence on CVA, especially when initial margin (IM) is also present.

What is margin settlement risk and why is it so important for CVA

Alexander Sokol, CompatibL
Alexander Sokol, CompatibL

Under a zero-threshold credit support annex (CSA) mandated by recent regulations, the parties in a bilateral trading relationship exchange variation margin (VM) to offset the exposure to each other. If they are also subject to the Basel Committee on Banking Supervision and International Organization of Securities Commissions IM regulation, IM is posted in addition to VM. The stated objective of adding IM to VM is to eliminate exposure and CVA almost completely.

As it turns out, the mechanics of CSA only work well away from trade payments. In their vicinity, it grinds to a halt in the face of new risk that CSA is ill-suited to handle in its current form: margin settlement risk. This new type of risk arises because of the peculiar way trade and margin payments are exchanged. Each trade payment changes portfolio value, instantly creating exposure. Due to the time needed to perform valuation and issue a call for collateral, the reciprocal margin payment that reduces exposure back to its baseline level is scheduled to arrive only a day or two later. If this date falls within the margin period of risk (MPR), the margin payment never arrives, resulting in counterparty credit loss in the amount of trade payment, an amount that is one or two orders of magnitude greater than baseline exposure.

The exposure spikes that appear because of this effect are very tall, but also narrow, extending only for the length of MPR – usually 10 business days – around each trade payment. For a large portfolio, the aggregate effect of a large number of exposure spikes is a significant increase in exposure on almost every business day, with the amount varying from one day to the next. Andersen, Pykhtin and Sokol estimated that margin settlement risk is responsible for 15–25% of exposure without IM, and for up to 90–95% of the exposure with IM. [2, 3] Without margin settlement risk, CVA under IM would be suppressed to such an extent that it would be reasonable to ignore it; with margin settlement risk, it is only suppressed by a factor of around 5–10: certainly a significant reduction, but not sufficient to disregard the residual CVA completely. Exposure spikes are responsible for the majority of CVA that remains under IM.

The increase of CVA due to margin settlement risk is undesirable, but is it possible to avoid it? 

Andersen, Pykhtin and Sokol proposed a mechanism for eliminating exposure spikes and excess CVA under the traditional CSA framework. [2] They proposed a minor change to how portfolios are valued for the purposes of CSA. This makes the trade payment and the reciprocal margin payment fall on the same day, so they can be settled via a payment-versus-payment service such as CLS. An alternative way to eliminate margin settlement risk is the new SwapAgent service by LCH, which nets trade and margin payments. Unlike the former method, the LCH service can only be used for certain types of bilateral trades, however.

Until exposure spikes near trade payments can be eliminated, their contribution to CVA must be calculated. It may seem that such calculation would require valuing the portfolio on a daily grid, making it too slow for practical use. Fortunately, Andersen, Pykhtin and Sokol developed a fast and accurate numerical technique for calculating exposure spikes with daily resolution without daily portfolio revaluation. [1] Source code for implementation of this technique is available free of charge from for model validation purposes, and commercially as part of CompatibL Risk software.


  • LinkedIn  
  • Save this article
  • Print this page  

You need to sign in to use this feature. If you don’t have a account, please register for a trial.

Sign in
You are currently on corporate access.

To use this feature you will need an individual account. If you have one already please sign in.

Sign in.

Alternatively you can request an indvidual account here: