View source for Template:M intro isda on termination

{{drop|C|ommerce gives the}} lie to the idea that life is a [[zero-sum game]]. This was [[Adam Smith]]’s great insight: things need not be [[Thomas Hobbes|nasty, brutish and short]]. 

Each of us will only strike a bargain if we think, on our own terms, we’ll be better off as a result. That being so, there is no reason to ''end'' an ongoing business relationship: all being well, trade is an [[infinite game]]. If we are good enough at it, we can keep its positive feedback loop going, for the mutual betterment of everyone, indefinitely. ''Infinitely'', even. 

Therefore, we wish our relationships well and pray Godspeed for their long and fruity lives. Should the plums dangling from this or that branch shrivel; if things become more trouble than they’re worth, we can call time and bid our relationship a peaceful transition to the ultimate hereafter. But things do not always work out. 

Therefore, we pack our trunk with tools, implements and weapons with which, if we must, we can engineer a faster exit. There are a few different ways this can happen. While lawyers will happily rabbit on about these hypotheticals in the ''specific'', we do not talk about them in ''general'' terms enough. 

Below, JC comes over all over-analytical and counts the ''types'' of ways to put a commercial relationship in the earth.

====Customers and service providers====
{{drop|N|ow the great}} majority of [[financial contract]]s are between a “''provider''” on one side — a [[bank]], [[broker]] or [[dealer]] providing a “''service''”, broadly described: money outright, finance against an asset or a financial exposure — and a “''[[customer]]''” on the other who ''pays'' for that service. The customer is, as ever, king: the service exists for her exclusive benefit: the provider’s only wish is to manage its resources to best provide that service and extract a [[fee]], [[commission]] or economic [[rent]] by way of [[consideration]] for it. 

“Providers” are indifferent to how the instruments they serve perform. They do not mean to be “the other side” of the trade. They are, loosely, ''[[intermediaries]]''. [[Agent|Agents]]. They match risk-takers, collect a fee and wish the parties well without taking sides: they are “[[compassion]]ate”, not “[[empathy|empathetic]]”. As long as their customers remain in fine fettle, they should never need, much less want, to ''terminate'' their services. That is how they earn a crust. 

So expectations on either side of a service contract are different: the customer takes risk and retains the prerogative to go ''off'' risk as she sees fit, by paying the provider’s fees and whatever it needs to terminate the arrangements it made to provide the service in the first place: its “[[breakage costs]]”.

But [[Ceteris paribus|all else being equal]], the provider ''cannot'' exit without the customer’s permission. A fixed term [[financial contract]], binds a ''provider'' in a way it does not bind its ''customer''. 

But the customer’s financial prospects may darken. She may turn out not to be as good as her word. The regulatory environment may change, making the services harder or more costly to provide. The provider may, justifiably, want out.

Where it is no longer sure of its expected return, the provider must have a set of “weapons” it can use to get out of its contracts. These fall into a bunch of different categories, as we shall see:

====Categories of termination====
{{drop|P|ut these “termination}} scenarios” into three categories: “''without cause''”;<ref>You hear these described as “no-fault” terminations, but there is no ''fault'' in a termination brought about by unforeseen externalities, either.</ref> ''external events'' and ''counterparty failure''. 

Within “without cause” there is an odd category of “''pseudo''-termination” rights that regulated financial institutions must have but would never use and which, curiously, relate to their ''own'' solvency. 

Within “counterparty failure” — which we might also label “default” — we might break them down into ''performance failure'', ''indirect credit deterioration'', and ''[[Merger Without Assumption - ISDA Provision|ISDA in-jokes]]''. 

=====“Without cause”=====
{{drop|T|erminations “without cause”}} arise ''just because'' — no fault, no pressing need; just a gradual drifting apart of interests. As we grow in life, the things we value change. Passions of youth dampen, we tend more towards scabrous songs of experience than exuberant songs of innocence. Here we prescribe a notice period long enough to allow a customer to make alternative arrangements, but otherwise, we wish each other well and carry on our way. These will generally be “clean-up” rights. They will exist under framework contracts and will not impair in-flight services. A service provider must honour any specific Transaction terms already on the books before being allowed to move on. Without cause termination is there to clear out low-value and dormant clientry from the administrative record: there may be ongoing compliance or operational costs of maintaining inactive clients on the books.

======Pseudo-termination rights======
[[Dealer]]s sometimes do have rights to terminate Transactions on notice without cause,but these will typically be ''pseudo''-termination rights: where a regulated institution must have the power to terminate transactions for capital reasons, even though it never expects to use them in real life.<ref>See {{isdaprov|Automatic Early Termination}}, which is an extreme example of a pseudo termination right: in that it triggers automatically. Much more to say about that on the {{isdaprov|AET}} page.</ref> These are a marker of our incipient failure in the [[The Victory of Form over Substance|battle between substance and form]].

For example, a [[swap dealer]]’s right to terminate a customer’s [[synthetic equity swap]] position on (longish) notice. The dealer can thereby treat its exposure as a “short-term obligation” for capital purposes — because it ''could'' get out, if it ''wanted'' to— and this is enough to get optimised regulatory treatment. 

But a sound-minded dealer ''having'' such a termination right is a different and distant thing from it ever ''exercising'' it. It might be ''forced'' to, in the direst of stress circumstances (where its own survival was threatened) — but in that case, with the dealer teetering, most vigilant customers would be moving valuable positions away in any case.

=====“Termination Events”: regrettable, but no-one’s fault=====
{{drop|I|t is in}} the nature of uncertainty that unwanted things can happen that are no one’s ''fault'', but yet beyond anyone’s power to control or ''stop''. [[Force majeure]] and [[Change in law|changes in law]], [[tax]]ation and [[regulatory capital]] treatment can make the continued provision of a service uneconomic or impractical.

Under the ISDA framework these events are described as {{isdaprov|Termination Event}}s. Typically, they are measured {{isdaprov|Transaction}}-by-{{isdaprov|Transaction}}, so do not shut down ''all'' exposure under the Agreement; only under those {{isdaprov|Transaction}}s which are directly “[[Affected Transaction - ISDA Provision|affected]]”. 

Since no one is pointing fingers or swearing, there is generally more leeway for the parties to get heads together and explore workarounds and solutions to avoid termination. An eventual decision to terminate, while regretted, may well arrive during a time of relative psychological calm, ''sans'' malice, and probably even by agreement.

=====Additional Termination Events=====
The {{isdama}} allows, too, for customised Termination Events. While these are filed under “Termination Event” they tend to be “defaulty” in nature, so we will deal with them below.

=====“Default Events”: do we have a problem here?=====
{{drop|T|hen}} there are termination rights that flow from something ''wicked'' about your counterparty. These the {{isdaprov|Non-defaulting Party}} will exercise unilaterally, without so much as a by-your-leave. Being a contract, the main category of “wicked facts” about your counterparty will be ''things it promised under the contract but did not do'': call these “performance failures”.

======Performance failures======
Direct performance failures — in old money, [[breach of contract|breaches of contract]] — tend not to be controversial: if you think “failing to do a certain something” is an unreasonable ground for terminating a contract, you should not agree to do it in the first place. 

The classic “performance failure” in a [[financial contract]] is a [[failure to pay]]. There could hardly be an obligation more “of the essence” of a [[financial contract]] than rendering in full ''what'' you promised, ''when'' you promised it. 

As a result there is little negotiation of payment failure default events. Nor, when it comes to it, is there much room for doubt as to whether one has happened. The payment either settled or it didn’t. Ops can tell you a few minutes after the cut-off time.

Ninety-five per cent of all ISDA close-outs — yes, I ''did'' just make that number up out of thin air, but I challenge you to disprove it — are triggered by a {{isdaprov|Failure to Pay or Deliver}}. It rather makes you wonder why we waste so much time haggling over the rest of the Events of Default.<ref>I put it down to “[[Parkinson’s law of triviality]]”: people spend most of their time attending to complicated things that don't matter, because they are take time and make us feel like we have achieved something. See also [[Qix]]</ref>

[[Parkinson’s law of triviality]] exemplified: there is an inverse relationship between ''how long you will have to argue'' about a given close-out right during onboarding and ''how likely you are to ever have to use'' it afterwards.

Other direct “performance failures” are a bit more oblique. Technical non-performance measures that are less “essential” than payment failure tend to have longer cure periods: whereas payment failure might have a one-day<ref>Three days under the {{1992ma}}.</ref> [[grace period]], you might have as long as 30 days to put right second-order breaches of an agreement before they become outright {{isdaprov|Events of Default}}.<ref>This makes them useless in practice.</ref> 

Performance failures that imply bad faith or moral shortcomings, like [[repudiation]], tend not to have [[grace period]]s at all, but seeing as it is quite hard to goad a counterparty into committing one of these, this does not make them any more useful. 

By order of how quickly you can accelerate them, then, here are the {{isdama}}’s “performance failure” {{isdaprov|Events of Default}}:
{{small|80}}
{{tabletopflex|100}}
|+ Itchy Trigger Finger Guide
{{aligntop}}
! Event of Default !!  Section !! Grace period 
{{aligntop}}
| {{isdaprov|Repudiation of Agreement}} || {{isdaprov|5(a)(ii)}}(2) ({{isdaprov|Defaulting Party}}) or {{isdaprov|5(a)(iii)}}(3) ({{isdaprov|Credit Support Provider}}) || None. 
{{aligntop}}
| {{isdaprov|Misrepresentation}} || {{isdaprov|5(a)(iv)}} || None. 
{{aligntop}}
| {{isdaprov|Credit Support Default}} (Total failure) || {{isdaprov|5(a)(iii)}}(2)|| None.
{{aligntop}}
| {{isdaprov|Failure to Pay or Deliver}} || {{isdaprov|5(a)(i)}} ||One {{isdaprov|Local Business Day}} after due date.
{{aligntop}}
| {{isdaprov|Breach of Agreement}} || {{isdaprov|5(a)(ii)}}(1)|| 30 days after notice of default.
{{aligntop}}
| {{isdaprov|Credit Support Default}} (direct default) || {{isdaprov|5(a)(iii)}}(1)|| Expiry of [[grace period]] in {{isdaprov|Credit Support Document}}.
|} 
</div>

======“Credit deterioration” events======
{{drop|D|epending on how}} you look at them, there are between three and four ''indirect'' {{isdaprov|Events of Default}}, that do not require the {{isdaprov|Defaulting Party}}’s performance failure directly under the {{isdama}}. These events arise from ''independent'' indications that it might be about to go ''[[titten hoch]]''. These are more fraught, because evidence for them may not be available, if they can be determined at all.

These are the “credit deterioration” events: {{isdaprov|Default Under Specified Transaction}} — where the {{isdaprov|Defaulting Party}} has defaulted to the {{isdaprov|Non-defaulting Party}} specifically, only under a separate master trading agreement (such as a stock lending agreement or a repo); {{isdaprov|Cross Default}} — where it has defaulted to someone else altogether under an unrelated [[Borrowed money|borrowing arrangement]] — and {{isdaprov|Bankruptcy}}, where the {{isdaprov|Defaulting Party}} crosses that [[phase transition]] into formal resolution or protection. It didn’t specifically fail under your {{isdama}} but the overwhelming odds are it is about to.

These events have great potential for the intervention of sod’s law. {{isdaprov|Cross Default}} particularly — JC has a long and overblown article about that — but there are aspects of the {{isdaprov|Bankruptcy}} (especially {{isdaprov|Automatic Early Termination}}) that have the potential for severe unintended consequences.

How are you meant to know how great is your customer’s private [[indebtedness]], much less that it has categorically defaulted on it? What if a lender granted a waiver or amendment? Is it still a default?<ref>In JC’s view: no.</ref>

These sorts of [[doubt]]s will wrack the credit department. Even senior credit officers will be cowed. A silvery [[SRM]], reeking of [[Evyan skin cream and L’Air du Temps|Evyan skin cream and ''L’Air du Temps'']], will storm in, hotly denying ''his'' [[platinum client]] would ever default, and that it is sacrilegious to even say so. He will have roped in any number of fragrant senior colleagues from wealth management to weigh in on his client’s behalf. Any firm not populated in its senior echelons by lizards will, therefore, sit on its hands. (This excludes [[Goldman]], which ''is'' populated in its senior echelons by lizards.<ref>Heh, heh: I’m just kidding fellas. Goldman bankers are no more lizardy than any other investment bankers.</ref>)

=====Customised Additional Termination Events=====
The tailored {{isdaprov|Termination Event}}s your [[credit officer]] would bid you crowbar into the {{isdaprov|Schedule}} tend to have the character of customer-specific ''credit deterioration'' indicators: [[NAV trigger]]s, [[key man risk|key person event]]s, ratings downgrade triggers all suggest the customer may shortly fail under your contract, even if it hasn’t yet. These you will spend the ''most'' time haggling about, and by immutable laws of soddery, they will cause you the most bureaucratic fussing post-execution. For every [[NAV trigger]] you use, a hundred you will have to interrupt your day to waive. Actually, more than that. Credit ''never'' acts on NAV triggers.

This should tell us counterintuitive things: for [[dealer]]s, not to worry too much if a customer bridles at your favourite {{isdaprov|ATE}}; for customers, not to worry too much if your dealer won’t back down about it. We all have our roles in the great pantomime.

=====ISDA in-jokes=====
This leaves {{isdaprov|Merger Without Assumption}} — where, through a [[corporate action]] the {{isdaprov|Defaulting Party}} is transformed into a different legal entity which is somehow no longer bound by the {{isdama}} at all. You might call this a type of “credit deterioration” event, but you could also consider it a direct “performance failure” — an outright ontological ''denial'', almost — under the present contract.

As far as I can tell — and I checked this with the learned author of {{coc}}, and my young apprentice, and neither put me off the idea — {{isdaprov|Merger Without Assumption}} as articulated in the {{isdama}} is basically a ''practical joke''. A kind of documentary pun, stuck in there at a dark moment in the sacred [[Wording]] in which, to lighten the mood, someone lobbed in this harmless nonsense while fully-armoured ISDA knights wrangled canonical text through the medium of hand-to-hand fighting. 

There is never a time at which one could exercise a {{isdaprov|Merger Without Assumption}} {{isdaprov|Event of Default}}.

====Recap====
So, there are three broad means of terminating a financial contract: in order of controversy, on notice, on account of externalities, and following default. When negotiating these rights it is always worth bearing in mind the parties’ respective interests from the contract. An end user wants cheap and reliable exposure to risk. A provider wants cheap and reliable source of commission income. Neither, beyond beating down the other’s expectations about fees, wishes her counterparty ill. We can sand edges off potential rancour by appealing to [[in good faith and a commercially reasonable manner|good faith and commercial reasonableness]]: take a step back and it becomes apparent that is all anyone wants. The nightmarish hypotheticals of your own counsel with a grain of salt: worrying you into haggling suits her own expectations about fees income. 

And of the default events, remember everything comes down to a failure to pay, or a fear of a failure to pay. Those representations, downgrade triggers and key person terms only matter at all to the extent they lead to, or forewarn, a failure to pay.