Set-off

Regulatory Capital Anatomy™ The JC’s untutored thoughts on how bank capital works. →This page on the premium JC Varieties of offset in financial services Criteria Settlement netting Set-off Close-out netting Friendliness Chummy Hostile All-out thermonuclear war Examples Settling up after a boozy weekend away. Optimising BAU cashflows in back office. Recouping amounts owed from delinquent debtor you happen to owe things to Closing out against bankrupt hedge fund Consent required Yes. No. No. In scope transactions As agreed by all. As selected by setter-offer All transactions under specific master agreement Due date Same date only. Same date only. Any (we will make them all the same). Currency Same currency only. Same currency only. Any (we will make them all the same). Capital Requirements Regulation: 575/2013/ (EUR Lex) | BIS publications: BIS capital requirements for bank exposures to central counterparties | leverage ratio | RWA | portfolio risk | systemic riskIndex: Click ᐅ to expand: Regulatory Capital Tell me more Sign up for our newsletter — or just get in touch: for ½ a weekly 🍺 you get to consult JC. Ask about it here.

Set-off
/sɛt ɒf/ (n., v.)

To square up. At its simplest, there is a “set-off” between cross-claims for money in the same currency owed on the same date between a creditor and a debtor. To set-off is to cancel out to the extent of the mutual debt: to discharge the smaller claim in full, and leave just the remaining balance of the larger claim outstanding.

Net-net, there are three concepts to bear in mind: close-out netting, settlement netting, and set-off. Related, but different things. Seeing how important each of them is to the opimisation of capital ratios, a matter of critical importance.

This page is about the general concept of set-off. There is a special page reserved for chat about the Set-off clause (Section 6(f)) of the 2002 ISDA. For reasons explored at that page, there was not a set-off in the earlier versions of the ISDA Master Agreement, though frequently, people put one in.

Easy, tiger.

You do not exercise a set-off right willy nilly. Unless you are, mutually, settlement netting (this happens a lot less in practice than you might thing), set-off is a drastic unilateral remedy which indicates a lack of trust in your counterparty and will be seen as enemy action.

You would not do it, without agreement, to any client you expected to keep.

So, generally, keep your powder dry. Use set-off as a remedy only following an event of default.

See, for example, Set-off in the 2002 ISDA under Section 6(f). So you don’t just do it for the hell of it.

Set-off and netting

Set-off and subrogation under a guarantee

A debtor cannot set off a subrogated claim against liabilities the guarantor has to that debtor. Would the converse situation apply? Could a debtor set off a subrogated claim by the guarantor against another liability owed to the debtor by the beneficiary of the guarantee? On one hand the set-off should have been applied before the guarantee has been called upon. On the other hand, what if the guarantee is expressed to be payable regardless of any set-off (as usually it would be).

The varieties and mysteries of set-off

Contractual set-off

Where each party to a transaction owes the other they may agree that, instead of making separate payments, the party due to make the larger payment should simply pay the difference. Set off provisions in the ISDA Master Agreement and 2010 GMSLA tend to go a lot wider, and allow (on default) a non-defaulting party to offset amounts owing against any liabilities of any kind owed by the defaulting party. This is clearly a drastic step and ordinarily would only be exercised as an utter last resort. See:

• Set-off (GMSLA)
• Set-off (ISDA)

Insolvency set-off

You can’t contract out of insolvency set-off under English law. As we have said elsewhere, bankruptcy is a strange time. The usual cast-iron certainties that accopmpany dealings amongst merchants in the financial markets give way to a spooky, primordial dreamtime, where nothing is sure. All are beset by phantoms, nightmares and the dark pantomime of the supernatural.

Lawyers do not like this, but it keeps them in employment, so do not shed too many tears.

Normally, as JC’s article about set-off describes, peace-time set-off is a sensible, self-help remedy where parties can take matters into their own hands, apply a quid against a quo of equal value, and no part of the jurisprudence of fairness will be upset. I owe you, you owe me, to save forking over equal sums, we call it quits. My claim for £100 from you is the same my liability for £100 to you, so we put the two together and they vanish in a puff of legal and mathematical logic. Simple.

Bankruptcy changes all this. When, as a going concern, you face a bankrupt, your claim for £100 is not worth £100. It may be worth nothing. You must take your place in the queue of creditors and find out. And if, at the same time, you have a liability to that very same bankrupt for the very same £100 it is still worth £100. The bankrupt can make you pay the whole lot. This seems unfair, but this is the game of cosmic chicken we play by extending credit.

So, you see, an enforceable set-off is a highly attractive proposition — for you. For the bankrupt’s administrator, on the other hand, at least as far as the bankrupt’s other creditors are concerned, avoiding that set off is a highly attractive proposition.

This is the palaver that close-out netting and the vaunted Single Agreement concept addresses. The trick it tries to pull is to work like an enforceable set-off, without actually being a set-off, precisely because in bankruptcy, set-offs are notoriously unreliable.

We have seen wishful game-playing in insolvency scenarios. This is unlikely to work, but the fact that it has occurred to people to try it will give you a glimpse into the venal minds that operate in the capital markets.

A large Texan Energy conglomerate has gone titten hoch, leaving behind it, in that smoking crater, a portfolio of energy swaps it bought from you — Debtor A — which, against all the odds, are heavily in the money to the bankrupt. You owe, let’s say, a couple of hundred million. Reports in the market suggest that a number of your competitors — Creditors B, C and D — are in the opposite position — they are all owed a lot of money and the bankrupt, being a blackened stump in the middle of said crater, is not going to be able to pay them. Their combined claims against the bankrupt add up to roughly $200m. Indications are that these contracts will pay 10 cents on the dollar. If your counterparties are lucky they’ll get $20m between them, and that will take 5 years and God knows how much time, resources and legal fees.

There is a trade to be done here: Debtor A buys the claims of Creditors B, C and D for $25 million dollars. The Creditors all avoid five years of the swamp, and get out for more money than they were expecting. This is a great outcome. Debtor A acquires $200m of notional claims against Bankrupt E for 12.5 cents on the dollar, and then sets off that £200m claim against the $200m it already owes Bankrupt E under the portfolio of energy swaps. Creditor A has paid $25m to avoid a certain £200m loss.

Now, let’s be clear: while bankruptcy laws differ wildly this would not work in a jurisdiction with sophisticated bankruptcy regime: though the bankrupt is not involved — so it would not be directly voidable, bankruptcy administrators tend to have wide discretionary powers, and courts generally will take a dim view of actions that usurp the function of the administrator, as this clearly would do. Whether it might hold up if engineered before a bankruptcy, however, is another question. It would make a great play.

Banker’s right to combine accounts

The banker’s right to combine accounts arises where a customer has multiple bank accounts, usually where some are in debit and some in credit. This stands to reason, obviously — unless there are multiple accounts, all overdrawn, but only some of them benefit from a security interest — though see also the rules about bailment. It is sometimes called a Banker’s set-off but, in 1972’s (still leading) case National Westminster Bank Ltd v Halesowen, Lord Buckley noted it isn’t a set off right so much as a function of accounting:

“Nor is it a set-off situation, which postulates mutual but independent obligations between the two parties. It is an accounting situation, in which the existence and amount of one party’s liability to the other can only be ascertained by discovering the ultimate balance of their mutual dealing.”

Equitable set-off

This is a self-help remedy available to a debtor whose cross-claim arises from the same transaction (or a closely related transaction) as the original debt. Under this device a debtor simply deducts its claim from the debt it owes and tenders any balance to the creditor. The sums in question must be due or, if representing unliquidated damages, a good faith and reasonable assessment of the loss. (Contrast with set off at law, which requires the claims to be determined by judgment of the courts).

In Geldof Metaalconstructie v Carves [2010] EWCA Civ 667 the leading judgment confirmed the equitable test as being “whether the cross-claim is ... so closely connected with the claimant’s demand that it would be manifestly unjust to allow it to enforce payment without taking into account the cross-claim.”

Cross-affiliate set-off

Not a thing: see triangular set-off (which is also not a thing).

Chicken Licken’s Guide to things that might defeat set-off

===Assignment and its effect on Netting and Set-off=== Could a right to assign by way of security upset close-out netting such that one should forbid parties making assignments by way of security of their rights under a master netting agreement (such as an ISDA Master Agreement or a 2010 GMSLA), for fear of undermining your carefully organised netting opinions?

Generally: No.

• An assignment by way of security is a preferred claim in the assignor’s insolvency over the realised value of certain rights the assignor holds against its counterparty. It is not a direct transfer of those rights to an assignee: the counterparty is still obliged to the assignor, not the assignee, and any claim the assignee would have against the counterparty would only be by way of subrogation of the assignor’s claim, should the assignor have imploded in the meantime or something.
• “Nemo dat quod non habet”: the unaffected counterparty’s rights cannot be improved (or worsened) by assignment and, it being a single agreement, on termination of the agreement the assignee’s claim is to the termination amount determined under the Agreement, which involves terminating all transactions and determining the aggregate mark-to-market and applying close-out netting. No one can give what they do not have.
• The assignee can be in no better position than the assignor and this takes subject to any set-off. The conduct of the debtor vis a vis the assignee is irrelevant, unless it gives rise to an estoppel. See Bibby Factors Northwest Ltd v HFD Ltd (paragraphs 38 and 48).

At the point of closeout, the assignee’s right is to any termination payment payable to the Counterparty. Therefore any assignment of rights is logically subject to the netting, as opposed to potentially destructive of it.

But: This is only true insofar as your netting agreement does not actively do something crazy, like disapplying netting of receivables which have been subject to an assignment and dividing these amounts off as "excluded termination amounts not subject to netting".

I know what you are thinking. "But why on God’s green earth would anyone do that?" This is a question you might pose to the FIA’s crack drafting squad™, who confabulated the FIA’s Professional Client Agreement, which does exactly that.

Collateral and set-off

Could your posting me collateral, or granting me a security interest, upset a set-off arrangement I have elsewhere?

Talk about first world problems. Who ever heard of too much credit mitigation? It seems absurd, but the JC has heard it cited, by a patient attorney as a risk, in this case:

A certain principal has entered transactions with you under collateralised agency master agreement. Say an agent lending agreement, or an agency ISDA Master Agreement. The Agent, for your behalf and those of your fellow principals, is relying on contractual netting to manage the whole exposure across its portfolio with you. If you go bust, the portfolio atomises into groups of transactions with individual principals which net down, but there may be some counter-indemnity or other mechanism whereby the agent may apply any excess collateral resulting from closeout against one such principal against a shortfall arising following close-out against another.

Now what happens if, independently of that agency arrangement, you grant security, or post collateral to that party? Does this defeat netting or set off?

Friends, it is hard to see how. If it is a pledge or security interest, and there is a shortfall on your portfolio, this is extra ballast, should you want it, but it hardly can interfere with the agent’s rights should it want to apply other collateral excesses to your shortfall. And if you have an excess — well, what is there to set off? If it is pledged by title transfer this might (if it applies) adjust the calculation of insolvency set-off, but surely only in a way that leads to your exposure, such as it is, being better covered rather than worse. Again, that should mean your agent has less, not more, need to dip into other pools of excess collateral to net out.

See also

• Set-off (ISDA)
• Set-off (2010 GMSLA)
• There is no general right of set-off in a 2010 GMRA
• For the net-net valuation technique, go read Graham on Security Analysis, and in no other circumstances utter that expression.
• Netting
• Triangular set-off