Template:M summ 2018 CSD 13(h)

What an omnishambles. ISDA’s crack drafting squad™ may usually be tiresome, leaden in its literary style, and pernickety to the point of distraction, but one thing you can say for it is that it does, usually, do things properly. It is thorough. It leaves no stone unturned, even when you wish it rather had.

With this provision it looks like the ’squad got to the point of maximum disarray, with all rocks upturned and slaters, bugs and cockroaches scuttling everywhere, and it just had a tantrum and stormed off. These provisions don’t even make sense. They are not even grammatical.

The basic problem, part I

The problem to be solved is this: initial margin is designed to cover mark-to-market exposure between (usually daily) variation margin calls. It is usually calculated to cover the likely possible change in portfolio value over that “liquidity period”, given the potential drop in collateral value over the same period. That is, one day.

However, when a counterparty goes titten hoch, the process of closing it out and determining who is owed what is a long process. For a big complex financial institution, can be months or years. One day’s market move starts to look a little bit meagre. Seeing as the initial margin is, by regulation, in the shape of non-cash assets, it too is subject to the vagaries of the market and can move up or down. So it might not quite cover what you thought it was going to cover.

to a great extent, that is just the non-linear unpredictable risk of the market. The answer is to take more collateral, of better quality, but that has its limits. So the alternative is to at least allow people in to convert the collateral into cash, to stop half of the portfolio moving around.

But fundamentally, this is just one of those risks it would be lovely to banish, but you can’t. Sorry, regulators!

The basic problem part II

Now remember: unlike variation margin, where only the in-the-money counterparty holds it, there are necessarily two buckets of Regulatory IM at all times: the stuff you posted as security for mark-to-market moves against you, and the stuff the other guy posted as mark-to-market movements against her.

Now: if a catastrophic event affects one party that precipitates a close-out, you stop exchanging variation margin. There’s no point: one side can’t pay it, Q.E.D.; the other side would be mad to pay it (and thanks to Section 2(a)(iii), doesn’t have to in any case).

At the last point that the parties exchanged VM, the net mark-to-market of the whole portfolio was (more or less) nil. After that point, until all Transactions are terminated, the MTM value of the portfolio will swing around. It could go either way. It does not follow that the Unaffected Party will be owed any money. By the time it has determined the Early Termination Amount, it may owe the defaulting party money. Until then it doesn’t need its own initial margin back, it should not get its initial margin back, and nor should it get to take the Affected Party’s initial margin.

This is just my opinion.

What was ISDA’s crack drafting squad™ trying to achieve?

So this brings us to the abomination we find on the page before us. God only knows what ISDA’s crack drafting squad™ thought they were trying to achieve. Whatever remote objective they had as a goal, and whatever contingencies were dogging the ’squad’s fevered subconscious as they trudged, in formation, through the moist, dengue-infested swamps of of this drafting exercise — and there is some talk that there may have been skirmishes with pockets of rogue buy-side advisors to distract them as they went waded through hip-high sludge — what is left to posterity is a confused, gibbering disaster.

What did they need to achieve? Straightforward.

All this provision does is describe when a Secured Party can actually take the initial margin the Custodian (IM) is holding for it — the return of its own initial margin, and the stuff the other guy has posted, assuming the other guy is the one who, at the end of the day, owes the money.

You should not be surprised to hear this should be, more or less, when the Chargor has actually defaulted and been closed out, the Early Termination Amount calculated, been found to be owed by the Chargor, and the Chargor having failed to pay it — and, really, the control of secured collateral held subject to a “Control Agreement” would ordinarily be most suitably dealt by that Control Agreement. The clue, surely, is in the name? Well, the 2018 English law IM CSD does its own job or determining when this would be — it does a horrible job of it, truth be told, but it is a job — so (anecdotally) the market-standard Control Agreements all tend to defer to the Secured Party Rights Event as determined under the 2018 English law IM CSD. So here we are.

What did they achieve?

An unholy mess. The “unless otherwise specified” option allows a Secured Party to access collateral from the Designation of an Early Termination Date — an arbitrary date, in the future, at which time you do not know whether you are even owed anything. You don’t even know whether you are owed anything on the Early Termination Date for that matter.

You are able, by electing “Failure to Pay Early Termination Amount” to allow access to Collateral at the point where the Early Termination Amount has been determined and the party required to pay it has not done so — which we think is the appropriate time — but only if the Early Termination Amount resulted from an Event of Default. Not if it arose from a Section 5(b) Termination Event.

Or you can elect to let your Control Agreement govern.

What would the JC suggest?

If you can resist the urge to fire them at ISDA’s headquarters, you can damn the torpedoes and take the JC’s recommendation, as discussed below.