(The opinions expressed here are those of the author, a columnist for Reuters)
By Andy Home
LONDON, Oct 7 (Reuters) - It was almost exactly six years ago that Deutsche Bank cancelled 100,000 tonnes of aluminium sitting in London Metal Exchange (LME) warehouses in Detroit.
It was the largest cancellation of metal in the LME’s history and bloated the load-out queue at sheds operated by Metro from a benign 20 days to a monster 120 days.
So began a saga that splintered global aluminium pricing, bitterly divided the industry between producers and manufacturers and embroiled the LME in a media and regulatory fire storm.
Six years on and there is a sense that this chapter in LME history is now drawing to a close.
Physical premiums have returned to something close to normality, albeit with some residual effects from the infernal queue machine.
The LME has completed an overhaul of its physical delivery network that has gone way beyond simply killing off the queue model.
And now a U.S. judge has dismissed the last remaining legal action against Goldman Sachs, Glencore and JPMorgan and their respective Metro, Pacorini and Henry Bath warehousing companies.
Even Judge Katherine Forrest of the Southern District Court of New York concedes there may well be an appeal, but the failure of similar parallel legal action at the appeal stage in August suggests a high legal hurdle if the case is to proceed further.
That failed appeal at the second circuit directly influenced the most recent judgment.
Judge Forrest had previously dismissed the case brought by companies classified as “commercial end users” and “consumer end users” of aluminium.
The appeals court upheld her judgement by 3-0.
“The manner in which that appeal was briefed and argued, combined with the Second Circuit’s rationale in its decision, cast serious doubt on the viability of the remaining claims,” by companies categorised as “first level purchasers” of aluminium, according to Forrest.
Forrest had quite a lot to say about the issues arising from such overlapping of cases, which the legally curious can read about in a blog by Alison Frankel on this link: reut.rs/2dRpqR7
But at its core the appeals court ruling was that the negative impact on aluminium purchasers from an explosion in the physical U.S. Midwest premium due to what Judge Forrest termed the “alleged shenanigans” in LME warehousing amounted to “collateral damage”.
If anti-competitive behaviour occurred, and the emphasis was on the “if” in that statement, then it occurred in aluminium warehousing services.
“The physical aluminum market is a secondary locale in this drama,” according to Judge Forrest, particularly since none of the plaintiffs were directly involved in storing or trading metal in the LME warehouse system.
Any alleged injury resulting from the load-out queue at Detroit was “suffered down the distribution chain of a separate market, and was a purely incidental byproduct of the alleged scheme.”
That finding by the appeals court applied equally to the “first level purchasers” case, according to Forrest.
All of which may seem like legal nicety to many in the aluminium market who watched physical premiums go super nova in tandem with the lengthening load-out queues, first at Detroit and then at the Dutch port of Vlissingen.
But the legal argument does capture a core truth about the much-disputed interaction of queue and premium.
Thanks to Senator Carl Levin and his detailed cross-examination of Metro and Goldman executives about the Detroit queue back in November 2014, there is plenty of public detail about that original Deutsche cancellation and the other “merry-go-round” deals that followed.
And it is clear that when Metro offered Deutsche the original “merry-go-round” deal, whereby the bank could earn discounted storage if it moved its metal off-market to Metro sheds and then re-warranted them in Metro sheds, no-one was thinking about the potential impact on the Midwest premium.
Rather, it must have seemed a clever solution to every warehouse operator’s dilemma of how not to “lose” metal to competitors once it is cancelled.
After all, if Goldman really had been thinking about manipulating the premium, it would have bulked up with physical metal when it bought Metro in February 2012.
But it held zero physical aluminium at the time. And even though it did start incrementally building a position, it amounted to only 93,000 tonnes by the end of the year, less than Deutsche had just cancelled.
It seems that Goldman and everyone else only realised there was a linkage between queue and premium after Deutsche’s big bang moment.
And they and everyone else started adjusting their trading accordingly.
Which is how markets work.
Something happens and the market adjusts, a process of evolution that generates an unpredictable chain reaction over time.
Remember when premiums went through the roof at the start of 2014? That was not due to the Detroit queue mechanics but rather to a bank trying to capitalise on the sudden demand for a premium hedging tool and getting its exposure wrong.
By the end of that year the market had adjusted further to the point that banks, funds and physical players were trading the queue itself, a trade based on the mathematics of queue length and resulting storage costs.
It took a barrage of measures from the LME to kill off this new queue market.
And it has succeeded, notwithstanding the legacy load-out queue at Vlissingen, which is now incrementally decaying.
Metro has since been sold by Goldman and the LME has slapped it with a retroactive $10 million fine. JPMorgan has also sold Henry Bath, leaving only Glencore in the LME logistics space.
And the aluminium market has moved on, leaving the queue-premium dynamic as an interesting but ultimately unresolved historical footnote.
Judge Forrest herself perhaps best summed up the linkage in her dismissal of the first case.
“That the combined actions of traders and warehouses to maximise their profits negatively impacted downstream purchasers through a rise in the Midwest Premium is clear but (...) this was an unintended consequence of rational profit maximising behaviour rather than the product of conspiratorial design”.
Pending appeal, of course. (Editing by Ruth Pitchford)