LONDON (Reuters) - Investment fund Cobalt 27 boasts on its website that it holds 2,904.7 tonnes of physical cobalt “fully insured and stored in LME-certified warehouses in Europe and the U.S.”
Which is curious because the London Metal Exchange (LME) reports only 840 tonnes of cobalt sitting in its entire global warehouse network.
That, however, is metal that has been placed on warrant with the exchange. What Cobalt 27 owns is not on warrant. It could be warranted overnight. It is, after all, already sitting in one of the LME’s 559 registered warehouses.
But because it hasn’t been warranted, it doesn’t get counted as part of the LME’s daily inventory reports.
And because it hasn’t been warranted, it’s also subject to a much lower storage charge. Which is, of course, one of the main reasons it hasn’t been warranted.
Welcome to the world of shadow LME warehousing, the latest chapter in the exchange’s long-running storage saga.
LME stocks of all metals have fallen from over 7 million tonnes in 2013 to 1.7 million at the end of June.
Much of this has been down to the erosion of aluminium load-out queues at LME warehouses in Detroit and the Dutch port of Vlissingen. LME rule changes have accelerated the process.
The number of registered LME warehouse units had been falling in tandem with stocks, which is logical since warehouse operators don’t make money by running empty sheds.
Over the last year though, the number of registered units has increased from 536 to 559, even as total registered LME stocks have shrunk by another 25%.
The withdrawal of Engelhart’s 22-LME warehouse operation has benefited Kloosterboer, which picked up the European warehouses, and ISTIM, which bought Engelhart’s Singaporean subsidiary.
ISTIM continues to expand its LME warehousing footprint with 61 units, putting it in fourth place behind C. Steinweg with 164, Access World with 106 and P. Global Services with 66. Between them the four companies were storing 78% of LME-registered metal at the end of June.
And quite possibly a lot of unregistered metal as well.
The discrepancy between falling visible stocks and rising warehouse units delineates the growth of the shadow storage system of LME-certified warehouses with stocks of unregistered metal such as Cobalt 27’s physical hoard.
It’s also noteworthy that while total registered LME storage capacity has been declining over the last five years, it hasn’t been falling nearly as fast as stocks.
LME inventory has shrunk by two thirds since 2015, but storage capacity has contracted by just 23%.
What’s there is being used. Lots of unfilled space is not in the average warehouse company operating manual.
Rather than being the latest Machiavellian wheeze by LME warehouse operators to replace the discredited load-out queue operating model, this shadow storage system is simply an efficient market reaction to a distorted LME warehouse pricing landscape.
Although the exchange is continuing for another five years its freeze on rental fees and load-out costs, decades of aggressive price hikes have left a yawning chasm between the full cost of LME storage and anywhere else.
Rent for a tonne of on-warrant aluminium can be as high as 56 cents per day. It’ll cost you another $45 if you want to get it loaded out.
You can of course negotiate a discount with the warehouse operator but your metal may already be covered by a pre-existing rental deal between the operator and the person who put the metal onto warrant in the first place.
Storing the metal outside the LME system, where rent can be as low as 6 or 7 cents per day, suddenly seems very attractive. That’s why so much aluminium was logjammed in load-out queues a couple of years ago.
The quid pro quo for cheaper storage, though, is the loss of what the LME calls its “gold standard” of security and liquidity.
Metal acting as collateral for the stocks finance and repo trade is on warrant because it can, if circumstances require, be sold immediately.
Moving metal off LME warrant generates a spectrum of higher insurance and financing costs depending on where it ends up.
Stocks financiers have in the past mitigated these extra costs by storing metal in an LME-designated location and in units close to LME sheds.
But why be “close to” when you could be inside?
LME shadow storage is the answer, an evolving arbitrage between the cost of holding metal on LME warrant or off-warrant in the same warehouse. The rent on the latter can be as low as 10 cents per day.
Moreover, the owner of the metal reaps the same insurance and financing gains as if it were on warrant.
Bit of a no-brainer, isn’t it?
Which is why LME warehouses are seemingly thriving even while visible stocks are grinding relentlessly lower.
The LME itself accepts that shadow storage is an efficient
and thriving business. Nor is the exchange in the business of setting stock holding targets.
But its consultation document released Thursday reflects the concerns of many parts of the LME user base that what’s in the LME warehouse system should ideally not be just the amount needed for physical settlement of contracts.
True, at times of market tightness shadow metal is readily available for warranting but the price comes in the form of time-spread tightness, sometimes severe.
LME registered tin stocks have rebuilt from a meagre 740 tonnes in May to a current 5,990 tonnes but it took a cash premium of $295 per tonne to suck that metal into the LME warrant world.
Lower on-warrant stocks imply more frequent periods of time-spread volatility, which may or may not be reflective of physical market tightness.
Market transparency, meanwhile, is much diminished.
This is not just a problem for analysts trying to interpret LME stock movements that often appear random because the drivers are storage not market dynamics.
Physical buyers are left disadvantaged by the appearance of low LME inventory and the shadow reality they can’t see.
The broader market should welcome the LME’s proposals to try and encourage more metal back onto warrant and to extend its reporting net to cover shadow stocks.
Because the shadow LME storage market is not going to disappear any time soon. It’s simply too efficient a market model for both warehouse operators and stock financiers.
The opinions expressed here are those of the author, a columnist for Reuters.
Editing by Elaine Hardcastle