Netting vs closeout netting

saurabhpal49

New Member
Hi David,

In Gregory chapter 4 there are separate sections for "netting" and "close out netting "

Is there any difference between the 2 as the examples provided for each of them also seems identical

Thanks
 
@saurabhpal49 , netting and closeout netting are not exactly the same thing (even if the netting part is the same in both)
However, closeout is another mitigation , let me quote Gregory here:

" Netting: the ability to offset all transactions (both in an institution’s favor and against it) when a counterparty is in default.
Close-out: this allows the termination of all contracts between the insolvent and a solvent counterparty without waiting for the bankruptcy to be finalized (which can take many years) "

So a close-out netting is a combination of close-out and netting.

I hope it helps.
 

saurabhpal49

New Member
Hi Eltanariel,
Even in netting we are offsetting the transactions which is similar to closing it ( as in case of closeout)

And both of them is triggered in case the counterparty defaults ,so what's the difference?
 
Close-out allows you to terminate the deals before bankruptcy, but it doesn't mean you can use netting
Netting allows you to net all your trades, but it doesn't mean you can do it before bankruptcy.

Nowadays most of the ISDA master agreement includes both netting and close-out but there is really a distinction between close out and netting
 

David Harper CFA FRM

David Harper CFA FRM
Subscriber
@saurabhpal49 But I do not think payment netting is triggered by counterparty default. The difference is between payment netting and close-out netting. The classic example of payment netting is an interest rate swap where, at each (eg) six-month settlement (aka, exchange), the fixed "coupon" payment is exchanged for the floating "coupon" payment, but they are netted (as in payment netting). If the swap performs without either counterparty defaulting, every cash flow exchange will be payment netted. In contrast to close-out netting, which only occurs if there is a counterparty default. Technically, the FRM (going back to Jorion in FRM handbook) parses credit risk into pre-settlement versus settlement risk. Pre-settlement credit risk is what we what typically think of as credit risk; i.e., default on loan or by counterparty, credit deterioration. Settlement credit risk refers to the very short term risk in a cash flow or payment exchange, such as the swap exchange or an FX currency exchange. This distinction associates with the netting distinction:
  • payment netting is mitigation for settlement risk
  • close-out netting is mitigation for pre-settlement credit risk
But I also think an easy way to differentiate is that payment netting happens in a normally performing contract where neither party is defaulting, but close out netting is triggered only if there is a default. Thanks!
 

Jaskarn

Active Member
@David Harper CFA FRM
There is one line in your notes
"Netted positions are inherently more volatile than their underlying gross positions, which can create systemic risk."

Why is this so? Please explain with an example.

Thanks
 

David Harper CFA FRM

David Harper CFA FRM
Subscriber
Hi @Jaskarn Consider you have with a counterparty two trades:
  • Trade 1 MtM = +10.0; i.e., you have a gain on this contract and therefore credit exposure
  • Trade 2 MtM = -9.0; i.e., you are underwater here with credit exposure = Max(0, -10) = 0
Your gross exposure is 10 but your netted exposure is only 1.0. Say the Δ is only +10% to Trade 1, then:
  • Trade 1 MtM = + 11.00
  • Trade 2 MtM = -9.0
Gross exposure increases only 10% to 11.0, but netted exposure doubles to 2.0! Similarly, any increase in the Trade 2 impacts the netted exposure but has no impact on gross exposure. For exampe, say Trade 2 jumps to -6.0:
  • Trade 1 MtM = + 11.00
  • Trade 2 MtM = -6.0
Gross exposure is unchanged, but netted exposure quadruples to 4.0! I hope that illustrates, thanks,
 

Jaskarn

Active Member
@David Harper CFA FRM

Please explain below graph a bit more.

I understood from the graph relationship between Correlation and netting factor, as correlation decreases netting factor decreases which means more netting benefit.
Also, for any particular correlation, as netting set increases then netting factor will decrease.

But I could not see, maybe due to 3D graph being plotted in 2D space, relationships between Netting set and correlation.



1550464168105.png
 
Last edited:

David Harper CFA FRM

David Harper CFA FRM
Subscriber
Hi @Jaskarn

Gregory's Figure 7.22 graphs the following three axises:
  • Size of netting set
  • Correlation (this graph happens to be non-negative correlations only)
  • [Z-axis] Netting factor
I think the two key relationships are:
  • As correlations decreases, the netting factor decreases. As the netting factor is the ratio net_exposure/gross_exposure such that a lower value (i.e., nearer to zero) is maximum/better while a 1.0 is lack of benefit/worse, this is an classic instance of "more diversification is measured by lower correlation such that lower correlation bestows benefits, in this case a better netting factor."
  • Less visible (to me) is the fact that a higher Size of the netting set is associated with a lower netting factor, although this appears to be more true (less true) when the correlation is lower (higher); e.g. at the top of the chart, where correlation is near to 1.0, it makes almost no different what the size of the netting set is.
With respect to your question about Correlation versus "Size of netting set," personally I have trouble interpreting this comparison because I perceive, in a way, these to the Independent Variables (in a manner of speaking) and the Netting Factor to be the Dependent Variable. I think fundamentally this chart is trying to illustrate the joint influence of Correlation and Size of Netting Set on the Netting Factor. I suppose we can attempt to "flip this graph" such that we draw a 2D plane thru a given Netting factor and observe its intersection with the 3D, in order to draw inferences, but visually I personally do not "see" anything compelling in this regard (i.e., that isolates on these two variables). I hope that's helpful,
 
Top