U.S. markets close in 5 hours 3 minutes

COLUMN-LME wins battle but its warehouse campaign will be long: Andy Home

(The opinions expressed here are those of the author, a columnist for Reuters.)

By Andy Home

LONDON, Oct 15 (Reuters) - The London Metal Exchange (LME) has won an important battle in the campaign to overhaul its dysfunctional physical delivery system.

Britain's Court of Appeal last week overturned a previous High Court ruling that the LME's consultation on linking load-in and load-out rates at log-jammed warehouses was unlawful.

Russian aluminium producer Rusal, which brought the legal challenge against the LME, has huffed and puffed, threatening to apply directly to the Supreme Court. This, however, looks little more than defiant posturing, given that the Court of Appeal expressly denied it the right of further appeal.

The LME can now implement its proposed solution, dubbed LILO, albeit only in February 2015, which amounts to a 10-month delay from the original April 2014 deadline.

Never mind. It has been clear for some time that warehousers have already changed their operating models to comply with the rule, which will penalise those that take in more metal than they load out.

Just a shame then that the two aluminium load-out queues at the LME good-delivery locations of Detroit and Vlissingen have flexed out again over the past month and that a new one, for zinc, has emerged in New Orleans.

This underlines the limits of the LILO proposal, a mixture of short-term containment and long-term attrition.

The broader campaign to exert more control over the exchange's warehousing network is only now starting, with the pending deployment of potentially far heavier firepower.


LILO is intended to do pretty much what is says on the box. If a warehouse operator with a load-out queue longer than 50 calendar days loads in more metal than it loads out over a stipulated period, it will be required to lift its load-out rate over the following three-month period in line with a preset formula.

It's a mechanism for preventing warehouse operators from accumulating ever more stocks and building ever longer queues - a virtuous circle for the warehouser, a vicious circle for just about everyone else.

The proposal is already working.

Metro, owned by Goldman Sachs and the dominant LME warehouse operator in Detroit, took in just 1,785 tonnes of metal between April, when the LME first started publishing per-operator stocks figures, and end-September. It loaded out 462,000 tonnes over the same period.

Pacorini, owned by Glencore and exerting even greater logistical control over the Dutch port of Vlissingen, last received metal in any significant volume in June.

Its net load-out rate over the April-September 2014 period was 205,000 tonnes, proof that it too has been operating in compliance with the spirit of the LME's rule, even while implementation was itself log-jammed in the UK courts.

******************************************************* Graphic on LILO at Metro Detroit: http://link.reuters.com/wen23w Graphic on LILO at Pacorini Vlissingen: http://link.reuters.com/ten23w Graphic on Queue Length at Detroit, Vlissingen and New Orleans: http://link.reuters.com/hun23w *******************************************************


How, then, have the queues started growing again?

What the LME's LILO rule can't do is change the amount of metal that was already sitting in exchange sheds prior to implementation.

While new inflows have been stemmed, changes in the status of existing stocks can cause queues to lengthen if that metal is cancelled in preparation for physical drawdown. Or, as was the case during the aluminium squeeze in August, queues will shorten if metal awaiting load-out is placed back on warrant.

The load-out queue at Vlissingen, for example, dropped sharply from 743 days at the end of July to 599 days at the end of August.

That reflected the re-warranting of 234,000 tonnes of aluminium more than the 62,675 tonnes of departures.

The process at Vlissingen has since gone into reverse, 108,435 tonnes of net new cancellations causing the queue to flex back out to 625 days at the end of September. More heavy cancellations this month, 146,150 tonnes through the end of last week, are likely to extend the queue further.

There were similar heavy cancellations last month at Detroit. The LME doesn't provide per-operator cancellation rates, but given the dominance of Metro particularly when it comes to aluminium storage, it's a fair bet that the 140,300 tonnes of net new cancellations were why its aluminium load-out queue rebounded to 702 days at the end of September from 620 days at end-August.

(No such interpretation problems at Vlissingen, by the way. Only one other operator holds stocks there, Worldwide Warehousing Solutions, with a negligible 300 tonnes.)

Metro Detroit has other metals trapped behind the wall of aluminium, which means that waiting times for those too increased from 74 to 119 days last month.

A raid on Metro's zinc holdings in New Orleans is why a new 71-day "flash" queue sprung up last month at the southern U.S. port. The amount of "live" on-warrant metal at Metro plummeted from 76,425 tonnes to just 775 tonnes over the course of September.


Metro now looks maxed out in terms of potential for further queue extension.

Its on-warrant stocks of zinc, or any other metal, in New Orleans are now marginal, while its total cross-metal stocks in Detroit shrank to just 55,910 tonnes at the end of last month.

At Vlissingen, by contrast, as of today there are still 426,325 tonnes of non-cancelled aluminium, which could potentially join the queue at the exit door.

It's not, in other words, necessarily game over at the Dutch port, and it's certainly not game over in terms of the broader issue of load-out queues in the LME system.

That "flash" zinc queue at New Orleans should serve as a warning that even if warehouse operators continue limiting intake, they do not control cancellation rates.

There are large concentrations of both zinc and copper stocks in Pacorini sheds at New Orleans, either of which could experience the same sort of mass cancellation that led to Metro's "flash" queue last month.

Moreover, there is no guarantee that warehouse operators such as Metro and Pacorini will not at some future stage start attracting more metal.

The reason why it isn't happening right now is as much to do with market economics as with good behaviour.

The previous linkage between the length of the LME aluminium load-out queues and the level of physical premium has been broken.

The perception that Metro's activities at Detroit were driving premiums ever higher relative to the underlying LME basis price is what caused all the hullabaloo about the exchange's warehousing practices in the first place.

It is now widely accepted that aluminium premiums have generated their own self-maintaining momentum, reflecting a cocktail of strong demand, lower supply and restricted access to stocks, whether sitting in LME or in off-market sheds with cheaper rents.

Although load-out queues at both Detroit and Vlissingen have flexed out again, the rental revenue from neither would currently be sufficient for warehouse operators to compete for fresh units in the physical market.

But will that still hold true in a year's time or even two years' time? Because that's how long the queues are. Supposing the physical premium froth blows off in the intervening period to the point that it becomes profitable again for operators to compete for metal with physical buyers.

That would risk reigniting concerns about the interaction between LME warehouse behaviour and physical premiums.

The LME's LILO rule would only mitigate, not prevent, such renewed competition.


There is also the more fundamental question of whether a load-out queue of over 700 days is in any way acceptable, either for the LME itself or for its users.

The clearing of the legal hurdle thrown up by Rusal opens the door to a new swath of potential changes to the LME's warehousing system.

That includes capping the amount of money that warehouse operators can make from metal sitting in a load-out queue. Which is ironic, since Rusal's primary objection in its legal challenge was that the LME had not given due consideration to this option in the consultation process.

The LME's defence was that it had received consistent legal advice in the past that it was unable to do so without incurring the wrath of European Commission regulators. Rusal, and for that matter, every one else in the market knew so, the LME argued, meaning it didn't have to include a detailed discussion of that option in its 2013 consultation on LILO.

Might things have changed?

The LME has since undertaken a deep-dive review of the legal framework determining its relationship with warehouse operators. The review is thought to have re-examined core parts of that relationship such as whether the exchange really can't have any say in overall rent and load-out charges.

Chronic inflation in both rent and charges in recent years has left LME storage uncompetitive relative to off-market storage, which is why so much metal is queuing to leave the system.

There has also been a logistical review of the whole warehousing system, extraordinarily the first time the exchange has ever conducted such a fundamental exercise.

Both reviews have been stuck in a release queue by the legal niceties of how the LME should conduct its consultations.

Either or both could lead to far more extensive overhauls of the system than the LILO proposal, which was only intended to be a holding operation.

The broader campaign is only just starting.

(editing by Jane Baird)