COLUMN-Aluminium premium bubble over? Just don't tell the LME: Andy Home

(Repeats Oct. 6 item. The opinions expressed here are those of the author, a columnist for Reuters.)

By Andy Home

LONDON, Oct (HKSE: 3366-OL.HK - news) 6 (Reuters) - So is the aluminium premium bubble well and truly over?

It (Other OTC: ITGL - news) certainly feels that way.

Japanese buyers have just secured a premium of $90 per tonne over London Metal Exchange (LME) cash metal for their fourth-quarter shipments. [

As recently as the first quarter of this year, Japanese premiums (PREM-ALUM-JP) were at a record high of $425 per tonne.

Moreover, these Q4 2015 premiums are the lowest since the third quarter of 2009, marking a return to historical norms.

The scale of the collapse in Japanese premiums is partly down to specific local drivers but even that statement reflects a return to normality, where physical premiums don't move in global lock-step but rather mirror regional supply-demand dynamics.

U.S (Other OTC: UBGXF - news) . and European premiums are higher at around $155 and $120 per tonne respectively but have fallen a long, long way since the start of this year.

Good news for the LME itself, which has faced fierce criticism for its warehousing policies, perceived by many users as the root cause of the aluminium premium bubble.

Its increasingly aggressive reforms of its loading-out procedures have undoubtedly played a part in bursting the premium bubble, although market forces have arguably played an equal if not larger part.

The exchange is still working on even more radical solutions to its load-out queues. Is it becoming over-obsessed with a problem that has now to all intents and purposes been fixed?

******************************************************* Graphic on historical Japanese aluminium premiums: http://link.reuters.com/zer75w *******************************************************

ASIAN COLLAPSE

Japanese aluminium buyers took heavy collateral damage from the premium bubble.

There were, remember, no LME load-out queues for aluminium in the Asian region. The United States had Detroit and Europe had Vlissingen. Queues at both LME delivery locations were core but not exclusive drivers of the premium machine of 2012-2014.

But aluminium is a global market and producers could argue that if Japanese premiums didn't rise to match those elsewhere, the metal would simply be diverted into the higher-premium regions.

Now Japan seems to be at the forefront of falling premiums, largely because it is located so close to China's overspill pipe.

The Chinese domestic market is chronically oversupplied and exports of semi-manufactured aluminium products act as the release valve. They have mushroomed by over 30 percent so far this year to 2.8 million tonnes.

The question of how much of this export stream comprises real semis and how much "fake semis", metal transformed just enough to qualify for tax rebates, is vexing the global market.

But even real semis act to displace products demand, generating a surplus of commodity-grade metal.

Evidence of that surplus in Asia is not hard to find.

Japanese port stocks stand at just under 500,000 tonnes.

The historical norm was 200,000-300,000 tonnes, with occasional but only brief deviations from that range. Even (Taiwan OTC: 6436.TWO - news) at the height of the global financial crisis, Japanese port stocks peaked at only 375,000 tonnes (in February 2009).

Then there was last month's single-day delivery of almost 100,000 tonnes of aluminium into LME locations in Asia. Port Klang in Malaysia received 67,925 tonnes and Singapore 27,775 tonnes.

Neither has received such volumes of aluminium for several years and the "arrivals" have been linked to traders liquidating stocks with the associated fear that there may be more to come.

GLOBAL COLLAPSE

But Japanese premiums are not falling in isolation. And they certainly wouldn't have fallen so hard had not U.S. and European premiums already collapsed from their super-charged peaks of over $500 per tonne.

And although neither region is immune from the Chinese flood, the unwind of the bubble is at least partly down to the LME's assault on those two load-out queues at Detroit and Vlissingen.

The exact linkage between queues and aluminium premiums remains hotly disputed and may never be fully resolved.

But the simple fact is that at one stage of the game, no-one bidding for free metal in North America could do so without calculating the waiting time at Detroit for the simple reason that Metro (Other OTC: MTRAF - news) , the warehouse operator that "owned" the queue, could itself bid at numbers matching the rental revenue reaped from the metal stuck in the queue.

The linkage broke down at the height of the bubble and has broken again during the bursting of the bubble.

But it's impossible to understand fully the premium collapse without factoring in the LME's determination to reduce its load-out queues.

This has taken the form of forcing queue-affected operators such as Metro and Pacorini Metals at Vlissingen to load out more than they load in, a formula that was tweaked again in August to force even faster queue decay.

As a result, Vlissingen loaded out 100,000 tonnes of aluminium last month and Detroit 56,500 tonnes.

Most of this metal has in all likelihood gone into alternative off-market storage but the steady stream of metal out of the LME system has undoubtedly played a part in the premium collapse, if only because the LME's determination to kill the queues has changed market behaviour.

PROBLEM SOLVED?

The collapse in premiums has also returned the global aluminium price structure to something close to the status quo ante.

As premiums ballooned, so did the gap between the LME price and the price paid by aluminium users, fracturing the "all-in" price in ways that undermined the validity of the LME contract.

That particular threat now appears to have receded, although the LME is still introducing aluminium premium contracts to compete with those already launched by CME.

It is also still tinkering with its load-out rules, which are becoming ever more complex and convoluted.

It is now proposing an "anti-abuse" mechanism to complement its proposal to cap rent for metal stuck in a queue. This is to prevent warehousers themselves being gamed by mass cancellations of metal and the formation of "flash" queues.

The answer, the LME thinks, is to stagger large cancellations of metal across several days but consultation on that means deferring yet another measure, a further lifting of the minimum load-out requirements, to March of next year.

Such continuous tampering with the rule-book risks falling foul of the law of unintended consequences, of which there are potentially many as the LME is finding out.

So should it desist from further reform now that both premiums and aluminium pricing structure have returned to something close to pre-queue normality?

No.

Firstly, there is a matter of principle here. The LME is surely right to ensure its warehousing system can deliver metal in a timely fashion, currently defined as 50 days to match a realistic sourcing and delivery timeframe for physical buyers.

It's simply wrong that anyone using a commodity exchange for physical delivery should have to wait years for that metal to become available.

Resolving the load-out pressures will not cure other broader problems with the LME warehousing network but it is a vital first step. And frankly, even if the LME didn't want to do it, it would probably be forced into action by regulators on both sides of the Atlantic.

Secondly, the aluminium premium problem may look as if it's gone but even now there is the potential for it to flare up again.

As of the end of August, the aluminium load-out queue at Vlissingen was 297 days and that at Detroit 254 days. Converting that waiting time into rental revenue and factoring in the load-out charge means the "value" of the queue at the Dutch port was around $200 per tonne and that at Detroit $160 per tonne.

Evidently, if either Metro or Pacorini were so minded, the old queue-premium arbitrage is back in the money and they could in theory re-enter the market to incentivise more metal into their sheds.

The fact that neither appears so minded right now is primarily down both to the LME's existing measures and its promise of more to come.

So, even if the premium bubble is now well and truly over and all the associated pricing problems are starting to look resolved, don't tell the LME.

It still hasn't fully tamed the existing queues and is looking to mitigate the potential for future queues.

Rightly so.

Because it's not just premiums that are reverting to the norm. The whole aluminium market may be reverting to its historical norm, one of chronic oversupply where manufacturers must compete with financiers for physical metal and where storage is the battleground.

Which, of course, was where the whole sorry story of queues and premiums originally began.

(Editing by Dale Hudson)