Podcast: McClelland on why you need a good MVA model

Numerix quant presents a model aimed at showing the total cost of a trade

Mauro Cesa

In this episode of Quantcast, our guest is Andrew McClelland, director in the quantitative research team at Numerix in New York.

He co-authored MVA, future IM for client trades and dynamic hedging, one of the two papers published in the August issue of Risk, with Serguei Issakov, senior vice-president in the quant research group at Numerix, and Alexandre Antonov, chief analyst at Danske Bank in Copenhagen.

McClelland talks to us about why it is important for a bank to have an accurate model to estimate the value of margin valuation adjustments (MVA), the cost of funding initial margin (IM). He explains that MVA “is a cost of doing business, a cost of trading, so it is natural that banks are seeking to reflect those costs in their pricing and their valuation”, similar to what happens with other XVAs, such as credit valuation adjustment and funding valuation adjustment. Ignoring it may lead to significant mispricing.

Their paper builds on research published in Risk in 2018, in which the same authors simulate and forecast the required sensitivities of future trade values using algorithmic differentiation.

Here, in response to requests from clients to have a model able to provide the total cost of a trade, including costs of hedging and funding, they extend the previous model to one that takes MVA into account both for the client side and the hedging side of the trade. “If you ignore the hedge side you are ignoring a cost,” McClelland points out.

They apply it to Bermudan swaptions, a suitable financial instrument for showing the performance of their approach.

The conversation then turns to the debate on whether funding costs should be included in the valuation of a trade. Interestingly, after years of intense exchanges of ideas between the two parties, the issue has not been settled yet. Some banks seem to take a more pragmatic approach and charge funding costs to their clients in a bid to make a profit, while others don’t charge them – and so gain a competitive advantage – if they can afford to omit them from the valuation.
 

Index:

00:00 Intro

01:40 Background on MVA and IM

04:33 Previous research

07:32 The proposed approach

12:34 Why a Bermudan swaption?

17:28 Computational hurdles

18:49 The risks of not computing MVA properly

20:50 The funding cost debate

27:28 Current and future research

To hear the full interview, listen in the player above, or download. Future podcasts in our Quantcast series will be uploaded to Risk.net. You can also visit the main page here to access all tracks, or go to the iTunes store or Google Podcasts to listen and subscribe.