2
$\begingroup$

Does the presence of a optional/mandatory right-to-break clause affect CVA calculations, and if so, how?

Given two (otherwise identical) 10y swaps with the same counterparty, one of which has a right to break at 5y (ours), intuitively I'd say the one with the break clause should have a lower CVA - if the counterparty default spread takes a dive we can exit the trade at replacement cost. The question is how do we take that into account when calculating CVA?

Definitions:

  • For the banks I deal with, when a break event occurs (regardless of whether the break was optional or mandatory) the trade is marked to market, a payment is made to whichever party is in the money & the trade is ripped up.

  • Trades with a mandatory break are often "replaced" with a similar trade. That might be because a counterparty isn't allowed to have a 20Y trade on their books, for example.

$\endgroup$
  • $\begingroup$ My current understanding is that the mandatory break case is relatively simple - the CVA on such a trade is equivalent to the CVA on a portfolio of an equivalent 5Y trade plus the expected value of the remainder of the trade at the 5Y point. The optional case is quite complicated however - there are many factors that might influence an institution into exercising an optional RTB - and that would be quite hard to simulate. $\endgroup$ – eddiewould Sep 25 '15 at 2:17
2
$\begingroup$

The presence of a mandatory break in a swap contract should reduce the CVA charge. That's because the CVA calculation models the default probability* swap market value while the swap is alive, so the calculation stops at the break date. There is one caveat: if a bank has a history of "waiving" mandatory breaks (i.e. In practice they never get exercised ) then the regulators can challenge his treatment. That would be a bank by bank discussion. So if you put one in, be prepared to execute it.

To clear up any confusion, a mandatory break is satisfied at the then market value, typically as determined by s panel of banks ( this needs to be specified in the confirm).

Optional breaks: if a counterparty grants a bank an optional break then then theoretically it should reduce the cva charge as for the mandatory breaks. However the implication is that on the break date either the break is exercised or the bank must demand a further cva charge to continue the swap (even if client credit is not impaired in any way). The cva desk may believe it will be unable to collect this value or even to execute the break against a valuable client. Hence the optional break may not achieve the full cva charge reduction at many banks.

$\endgroup$
  • $\begingroup$ That's pretty much as I thought. Now, what about optional break clauses? $\endgroup$ – eddiewould Nov 26 '15 at 0:57
1
$\begingroup$

would it affect CVA? that would depend on what happens on breaking.

Normally if we break a swap, the swap is over and there are no more cash-flows. We would only break if the NPV is negative and in that case we have no credit exposure so no effect.

However, CVA is generally computed on a book basis and that is more complicated. Ultimately, you have to simulate and PV the book on each default date. That PV will change if the swap has a break clause.

$\endgroup$
  • $\begingroup$ Wouldn't a swap with a positive NPV but where the counterparty has a poor/declining credit rating be a perfect candidate for breaking? I was under the impression that a break clause wouldn't affect the PV of a swap (ignoring counterparty risk) because the swap will be settled for replacement cost when broken. $\endgroup$ – eddiewould Aug 24 '15 at 5:26
  • $\begingroup$ normally swaps with break clauses cease to exist on breaking and no further cash-flows. $\endgroup$ – Mark Joshi Aug 24 '15 at 5:33
  • $\begingroup$ I've modified my question to state how break clauses work at the (Australian) banks I've dealt with - namely, when the break event occurs, a payment is made to whichever party is in the money (current value of trade) and the trade is ripped up. $\endgroup$ – eddiewould Sep 25 '15 at 2:14
-1
$\begingroup$

If you're calculating trade-level CVA, wouldn't it be the equivalent of calculating CVA as if the trade had a 5y maturity instead of 10y?

$\endgroup$
  • $\begingroup$ No - when the break event occurs, the trade is marked-to-market at that point and a settlement made to whichever party is in the money. Think of it this way - at the 5Y point (when the trade was due to be ripped up) we might be extremely in the money. If our counterparty has already gone belly-up by then, they won't be able to pay us the NPV of the remainder of the trade (which they otherwise would have). $\endgroup$ – eddiewould Sep 25 '15 at 2:12

Your Answer

By clicking “Post Your Answer”, you agree to our terms of service, privacy policy and cookie policy

Not the answer you're looking for? Browse other questions tagged or ask your own question.