Lehman Brothers International (Europe) v AG Financial Products, Inc.

From The Jolly Contrarian
Jump to navigation Jump to search
The Jolly Contrarian Law Reports
Our own, snippy, in-house court reporting service.
JCLR.png
A shelf in the JC’s law library yesterday
Editorial Board of the JCLR: Managing Editor: Lord Justice Cocklecarrot M.R. · General Editor: Sir Jerrold Baxter-Morley, K.C. · Principle witness: Mrs. Pinterman

Common law | Litigation | Contract | Tort |

Click ᐅ to expand:

Comments? Questions? Suggestions? Requests? Insults? We’d love to 📧 hear from you.
Sign up for our newsletter.

Lehman Brothers International (Europe) v AG Financial Products, Inc.

A case that came to the Supreme Court, New York County[1] and which concermed our old friends Lehman horcrux and the 1992 ISDA and, in particular, its Loss method of determining an Early Termination Amount.

While there is no Loss concept in the 2002 ISDA this case is relevant in that a 2002-style “Close-out Amount” is broadly the same thing as a Loss calculation.

The case concerns a portfolio of credit derivatives between LBIE and monoline insurer AG Financial Products, which the judgment confusingly labels “Assured” — AG once stood for “Assured Guaranty” — even though AG was writing credit protection, not benefiting from it.

In any case, LBIE created some tranched CLOs and mortgaged backed CDOs. AGFP wrote credit protection on the senior tranches. The CDOs “blew up” in the sense that every structured product in the world in 2008 blew up, as did LBIE and, nearly AGFP. LBIE having defaulted by Bankruptcy, AGFP (eventually) closed out its ISDA. As Non-Defaulting Party, AGFP was entitled to value its replacement cost and did so claimed it was owed $20m. LBIE said, “I’m sorry, you freaking what?” and countered that the close-out value of the contracts was $485m, payable to LBIE.

Now, one can and at times of stress should expect a difference of opinion in the value of unfunded derivative contracts. That s natural. But a difference of half a yard? That is a lot.

Cue litigation.

The first question, resolved in an earlier trial, was who had the burden of proving what. Crane J put this, as she put much in her judgment, drily:

When this trial first started, the court was skeptical about how, given such a large discrepancy, [AGFP]’s calculation could ever be reasonable. As the trial progressed, however, there was a growing realization that the marriage of catastrophic, historical financial circumstances with bespoke contractual terms, along with strong structural protections in the Securities, lent itself to the conclusion that at least LBIE’s calculations were not reasonable. ... Now, even with the benefit of that rebuttal case, and extensive post trial briefing, the court’s growing realization has crystallized to conclusion. LBIE’s valuation, that relied entirely on market prices its experts constructed for this litigation, was insufficient to meet LBIE’s burden to prove its own calculations were reasonable.

What can we learn from this:

Litigation takes ages

Here we are in the year of our Lord 2023, the Lehman horcrux was buried deep into the misty grags Iron Mountain some 15 years ago — that is two limitation periods — and we are still conducting litigation about how to close out and value an ISDA Master Agreement, a master agreement that presides over some trillions of financial instruments each year.

No-one really understands the practicalities of how ISDAs close out

This is, of course, because shots are not usually fired in the termination of an ISDA Master Agreement. And we suspect the sorts of instruments traded under ISDAs are, by and large a lot more vanilla and liquid than tranched CDOs.

When hair is on fire all bets are off

As we have written elsewhere, ISDA Master Agreement was drafted by committee in a time of peace, is negotiated in a time of prosperity and with hopeful expectation, and is often closed out in a time of apocalypse when it is not just the Defaulting Party who is preparing to meet its maker, but the Non-Defaulting Party and, as like as not, those putative Reference Market-makers too. Lions are lying down with lambs, plagues of locusts swarm the land and risk officers’ grasp of commercial and economic fundamentals — tenuous at the best of time — disconnects from reality altogether. There was a passing nod to our old friend the flawed asset clause, though it didn’t finally come up as a topic: Lehman entered Bankruptcy on September 15 2008, entitling AGFP to close out, but AGFP did not do so for ten months, on 23 July 2009. In the mean time we presume it suspended its payment obligations, as it is entitled to do under Section 2(a)(iii). It did this not just because it was allowed to, but from its own solvency perspective, it had no choice.

“Dealer polls” are not a thing

Oh, they are a legal contract thing, totally: they are just not a “this exists in the real world of derivatives trading” thing. The fact that so many legal contracts rely for their successful conclusion upon a dealer poll as a means of resolving valuationdisputes — the ISDA Master Agreement does, as does the 2010 GMSLA and the Global Master Repurchase Agreement — provision and no-one knows this, really only goes to show how pointless and falsely comforting is much of the verbiage that accompanies the construction and execution of modern financial instruments.

The Loss method

Rather than seeking dealer quotations, AGFP ran scenario analyses for defaults on the portfolios of mortgages it was insuring if they were allowed to run to term (which is when its obligation to pay any losses on the portfolio to insured tranches would come due). This was an audited, checked process and the court found it plausible and consistent with what you would expect an insurer to do. </ref>It wasn’t mentioned, in the judgment, but by the time the trial rolled around the CDOs would have long since matured so you would think they would know the exact default levels — but still.</ref>

LBIE’s rebuttal evidence, which rejected these actuarial methods and corroborating views in favour of projections obtained from other banks, the court felt, was a load of self-serving hooey. This is not a quote; just a general impression. “Suspect”, “quite dubious”, “not objective”, “had every incentive to make the outlook for the Transactions seem as dire as possible in order to set the stage to collect more from their monoline insurers” — these are direct quotes.

Thus, under the ISDA Master Agreement, the non-defaulting party “need not” consider market prices, especially where to do so would render the Termination amount “commercially unreasonable.” By limiting the use of market prices, ISDA thereby contemplates a situation, like the one here, where market prices are completely divorced from value. In addition, ISDA’s “Loss” provision is flexible enough to take into account all types of “loss of bargain,” even Assured’s, which had nothing to do with market prices.
Where reference to market prices does not always lead to a commercially reasonable result, courts favor a valuation that does not use market prices. ...
Market prices leading to a commercially unreasonable valuation is especially likely during periods of severe market disruption where reliable market prices may not exist.

The court didn’t believe the bank’s valuations. The quaint notion that a dealer poll would, at the point when needed, actually do anything was laid to rest in the 2023 case of Lehman Brothers International (Europe) v AG Financial Products, Inc. which involved the closeout and valuation of a 1992 ISDA following Lehman’s collapse.

This case is an object less on for many unacknowledged facts about derivatives trading — such as that cases involving seemingly tried and tested aspects of close-out methodology get litigated at all, let alone that they take 15 years to get to judgment — but the standout point is the forlorn pointlessness of convening dealer polls.

From Crane J’s factual summary:

In accordance with its responsibilities under the ISDA Master Agreement, following its declaration of an event of default, Assured engaged the assistance of Henderson Global Investors, Ltd. (Henderson), to conduct an auction so that it could satisfy the ISDA Market Quotation process. Henderson contacted 11 potential bidders in advance of the auction that took place on September 16, 2009. Not one bid was received.[2]

LBIE did manage to get some indicative bids that were, expert witnesses thought “indicative market data of where these transactions, these underlyings would be trading at that stage on termination date”. But not one of them was prepared to make a binding offer, and the most fulsome indicative bid was disclaimed up the wazoo:

“This is not investment advice of any kind and we do not purport any degree of accuracy in these levels.”

Useless, you would think, as an input in determining a fair market level. Indeed, internal LBIE emails — kids, if you learn one sodding lesson from the history of financial market disaster let it be “don’t put your darkest thoughts in emails to your buddies” — suggested they only wanted indicative bids to encourage other banks (many may have had similar trades on their books as LBIE), to make any bid, so that LBIE could then argue there was a market price:

“any color is good color to us and [JP Morgan employee] is lobbying for [JP Morgan’s US trading team] to at least put a number on it even if it is zero”.

Crane J notes, somewhat drily: “This raises the concern that LBIE’s goal, with respect to the indicative bids, was to make these trades seem as worthless as possible to then be able to collect the most from Assured in a lawsuit.”

So here are some things to bear in mind before reaching for a dealer poll to unblock a negotiation that is stuck on valuation:

Firstly, at the point you are likely to be arguing about it, everyone’s hair — yours, the counterparty’s and the rest of the market’s — hair will be on fire. Prices will be yoyoing around and most people will be focussed on their own book and won’t care about yours. Imagine your reference dealer is sitting on one of those mechanical bucking broncos. At the moment you ask for a firm bid on your portfolio — that, by the way, you don’t intend to hit — someone switched the bronco on full.

Secondly, since it has nothing to gain from providing a price — you want “price discovery”, not an actual trade, remember — no dealer in its right mind will give you one. Best case scenario it is distracted from managing its own book while bronco machine is on max. Less edifying ones are that it could get called as a witness in the litigation that is bound to follow or, God forbid, joined as a defendant in it. All your incautious bloomies are suddenly discoverable before an unsympathetic court.

See also

References

  1. This is a state, not federal, court: Civil Branch of the Supreme Court of the State of New York, New York County. Our court is the highest trial-level court for civil cases in the state court system in New York County
  2. Emphasis in original.