82,903
edits
Template:M intro csa (VM): Difference between revisions
no edit summary
Amwelladmin (talk | contribs) No edit summary |
Amwelladmin (talk | contribs) No edit summary |
||
| Line 1: | Line 1: | ||
{{drop|G|od bless them}}, when {{icds}} were deep in the ritualistic sacrifice phase of the [[Wording]] for the regulatory margin CSAs someone — we like to think it was [[Ser Jaramey Slizzard]] himself, but it is not documented — hit upon the idea of suffixing certain of the [[definitions]] in the [[Modern CSA]]s with “(VM)” — for “variation margin” — and (IM) — for “initial margin”. So assiduous were | {{drop|G|od bless them}}, when {{icds}} were deep in the ritualistic sacrifice phase of the [[Wording]] for the regulatory margin CSAs someone — we like to think it was [[Ser Jaramey Slizzard]] himself, but it is not documented — hit upon the idea of suffixing certain of the [[definitions]] in the [[Modern CSA]]s with “(VM)” — for “variation margin” — and (IM) — for “initial margin”. | ||
So assiduous were {{fkoi}} that the expression features no fewer than 235 times across the {{2016csa}} and a full 335 times across the {{2016nycsa}} scattered like buckshot across the posterior of a frolicking trespasser caught ''in flagrante delicto'' in his field by an angry farmer. | |||
Here is what 335 VMs look like in a job lot: | Here is what 335 VMs look like in a job lot: | ||
| Line 14: | Line 16: | ||
Said truculent child did as truculent children do: instead of accepting its punishment with good grace, understanding it has erred, and treating the whole incident as a teachable moment, it did its best to work around it. Swap transactions that were already on foot before the commencement of the new regulations were “[[grandfathered]]” and were not, specifically, required to be margined. Certain classes of transaction — not many, in the end — were out of scope for compulsory margin. So the industry prepared for a world where single {{isdama}} might have ''two'' CSAs: a regulatory one for in-scope products after the commencement date, and a non-regulatory one for grandfathered products and those out of scope after the commencement date. | Said truculent child did as truculent children do: instead of accepting its punishment with good grace, understanding it has erred, and treating the whole incident as a teachable moment, it did its best to work around it. Swap transactions that were already on foot before the commencement of the new regulations were “[[grandfathered]]” and were not, specifically, required to be margined. Certain classes of transaction — not many, in the end — were out of scope for compulsory margin. So the industry prepared for a world where single {{isdama}} might have ''two'' CSAs: a regulatory one for in-scope products after the commencement date, and a non-regulatory one for grandfathered products and those out of scope after the commencement date. | ||
Now running two CSAs on the same ISDA is a ''recipe'' for confusion. The whole point of a | What is the difference between regulatory and non-regulatory margin CSAs? Largely one of theory and not practice: a plain old [[Ancient CSA]] has a wide range of {{{{{1}}}|Eligible Credit Support}}, presumes you may set an {{{{{1}}}|Independent Amount}} (this is ISDAs traditional way of dealing with initial margin), and allows the parties to agree {{{{{1}}}|Valuation Date}}s more or less as they see fit and set often asymmetrical {{{{{1}}}|Thresholds}}, {{{{{1}}}|Minimum Transfer Amount}}s. | ||
The new VM CSAs (and more importantly, the regulations with which they were designed to comply) assume each Local Business Day would be a {{{{{1}}}|Valuation Date}} each day, that your {{{{{1}}}|Threshold}}s will be zero, and that {{{{{1}}}|Eligible Credit Support}} will take the form of {{{{{1}}}|Cash}} in the {{{{{1}}}|Base Currency}}. | |||
Now, in practice, old-style CSAs had in practice been converging on standard terms that were not so different from what the regulation required. Most were valued daily, with low Thresholds, and in cash. There might have been some advantages and balance-sheet efficiencies to be had from sticking to ''l’ancien regime'', but they were not as stark in practice as they look in theory. | |||
And operations people noticed that exchanging cash only is a whole lot easier, settles quicker, and is less volatile and easier to value than securities, but it is also not nearly as expensive as it seemed in principle: in client facilitation business, generally for every dollar posted out the door one comes in from someone else, so the net funding consequence of posting in cash was muted. | |||
Now running two CSAs on the same ISDA is a ''recipe'' for confusion. The whole point of a Credit Support Document is to be a single aggregating force, aggregating and boiling down all these diverse exposures and liabilities to a single number, and now there were two Credit Support Balances, two sets of Credit Support Eligibility Criteria, two Credit Support Balances and so on: in fact, ''twenty'' duplicated terms in the {{2016nycsa}} and its ancient equivalent (seventeen for the {{2016csa}}). | |||
You can imagine this could become very difficult for the poor bugger in ops that is handling two sets of collateral flows across 25,000 ISDA arrangements. | |||
{{icds}} elegant solution was to suffix the regulatory margin versions of these terms — what we call the “[[Modern CSA]]s” with “(VM)”. Result: a lot of (VM)s. ISDA’s prose in the CSAs — especially the [[Modern CSA]]s — is purple enough without these ugly parentheticals. | |||
But needed, right? Are we now in a world of parallel CSAs, carefully arbing the | |||