(The opinions expressed here are those of the author, a columnist for Reuters.)
By Andy Home
LONDON, March 31 (Reuters) - Japanese aluminium buyers will pay a premium for their second-quarter metal deliveries of $128 per tonne over the London Metal Exchange (LME) cash price.
This represents a sharp jump from the $95 negotiated for first-quarter shipments. It is also the highest premium since the second quarter of 2015, when the aluminium premium bubble was rapidly deflating.
Meanwhile LME stocks of aluminium sank below the 2 million-tonne level earlier this month and at a current 1,886,400 tonnes are at their lowest since December 2008. Excluding the metal that is awaiting load-out, the on-warrant total of 1,003,275 tonnes is the lowest since May 2008.
The LME outright price is trading just shy of the $2,000 per tonne level for the first time since early 2015.
It's a combination that in most commodity markets would indicate a tightening in availability.
But this is aluminium, a market where nothing is ever quite as it seems.
China, after all, is still lifting production and everyone knows that there are huge off-market stocks.
So just how tight is the aluminium market?
Graphic on Japan's quarterly premiums:
Japan's quarterly premiums, which act as a benchmark across much of the Asian region, are in part a reflection of local market conditions.
It's worth noting that regional supply has taken two hits recently.
The Portland smelter in Australia was knocked out by a power cut last December and is expected to take six months to fully recover. The Boyne smelter, also in Australia, is cutting output by around 14 percent due to rising power prices.
But local premiums are also reacting to what is happening on the other side of the world.
The U.S. Midwest premium is currently trading at just below 10 cents per lb ($220 per tonne) on the CME exchange, having bottomed out at around 6 cents per lb in the third quarter of last year.
Multiple domestic smelter closures over the last few years have left a widening supply gap in the U.S. market and imports are rising.
Imports of unwrought aluminium surged by 25 percent last year to 4.26 million tonnes.
Canada remains the country's biggest supplier but imports from Russia more than doubled to 722,000 tonnes and those from the United Arab Emirates surged by 87 percent to 547,000 tonnes.
A rising Midwest premium is rooted in the fact that the U.S. market is having to work harder to pull in more metal from further afield.
So far so good.
But do these fundamental drivers explain the scale of the current outflows from the LME storage system?
After several years of reform the LME warehouse system has been re-designed to prevent the sort of structural load-out queues that distorted the aluminium premium market over the 2012-2015 period.
This entails both metal leaving more efficiently than in the past and warehouse operators becoming wary of how much they take in lest they fall foul of the penalties stipulated for not loading out fast enough.
Throw in sporadic tightness across the nearby time-spreads and mass cancellations, almost 400,000 tonnes in February alone, and the push of metal out of the system is probably stronger than the pull of any physical market tightness.
At the end of February, for example, the residual load-out queue for aluminium at the Dutch port of Vlissingen (162 days) was complemented by "flash" load-out queues at both South Korea's Busan (136 days) and Malaysia's Port Klang (57 days).
Operators at all three locations (Access in Vlissingen and P Global Services at the other two) are incentivised to move metal out just as fast as they can, which helps explain why LME aluminium stocks have been falling at an average rate of almost 14,000 tonnes per day this month.
Annualised, that outflow is equivalent to around 5 million tonnes and no-one is forecasting anything like that sort of supply shortfall in the global aluminium market.
The inference is that much of this metal is simply moving to cheaper off-market storage, where it sits in the statistical shadowland.
That's not to say that the aluminium market isn't shifting towards a state of supply deficit, just to note that LME stocks are a poor way of quantifying any shortfall.
Commodity consultants CRU, for example, estimate that the supply deficit outside of China increased from 194,000 tonnes in 2015 to 821,000 tonnes last year. They are expecting it to balloon further to almost 1.3 million tonnes this year.
China's structural tendency towards supply surplus in part offsets this supply-demand gap everywhere else, although not last year, according to CRU, which posits a small deficit in China as well.
China's exports of aluminium in the form of semi-fabricated products are the world's balancing mechanism but one that is increasingly prone to disruption from trade disputes.
Nor has the country exported sufficient material to prevent the need for inventory drawdown elsewhere, although there is much devil in the detail of how much it has exported and, critically, in what form.
CRU calculates total global stocks, including what we can see on the world's exchanges and what we can't see in off-market inventory, will fall from 84 days worth of consumption in 2015 to 71 this year with a slightly faster attrition rate in the world outside China.
That's still a high level of inventory. And the problem is that those stocks are getting increasingly difficult to see as metal is hoarded in off-market storage away from the statistical light.
It is also problematic that at a time of rising outright price and rising premiums whoever is holding those stocks has no incentive to release them in any market-affecting way.
So is the aluminium market tightening? The answer is almost certainly yes, but not nearly as much as it might appear.
And that disconnect between appearance and reality is probably going to become more acute over time. (Editing by Greg Mahlich)