CAUTIONARY TALES
SLOW BOAT TO CHINA (TGITM Nov-Dec 2018) – Sunk
“Slow boat to China” discussed the (ab)use of LME “warehouse receipts” and “bills of lading” to obtain domestic Chinese financing, which has been a feature of the metals market since 2010 after the Chinese authorities attempted to cool their runaway property market by restricting bank loans to the sector.
In response, some property developers started pledging their copper imports to get cheap “collateralized” loans from domestic lenders, who didn’t secure the collateral.
The process could work as follows:
• Cheapest LME warrants were picked up through clearing, at the back of a warehouse queue (perhaps New Orleans, with a 6-9 month arrival time in Shanghai)
• The warrants were cancelled, generating warehouse receipts which were shown to a domestic Chinese lender
• Within a week or two, funding was provided for 6-9months, and the borrower dropped the warrants back into warehouse to avoid finance and storage costs for the rest of the period.
Realizing that the warehouse receipts weren’t retained by the lending banks, this also meant that the same warehouse receipts could be used to obtain loans from different banks simultaneously, with multiple pledging estimated to be 10x, 20x, even 50x any notional copper value in transit.
With China clamping down on the use of copper warrant financing in summer 2013, the practice switched to nickel and tin, with massive cancellations and re- warranting but very little metal leaving LME warehouses, i.e. just a process to generate warehouse receipts.
Qingdao scandal The Qingdao scandal in 2014 involving missing aluminium briefly threatened to expose the collateralized metal market in China, but authorities blocked access to the warehouses claiming it was an isolated issue.
After Qingdao, some banks and brokers suffered further losses over missing nickel and fake warehouse receipts. Despite this, domestic Chinese banks continued to lend to this “collateralised” funding market, suggesting that they were satisfied not to audit the collateral as long as they were repaid on time (possibly through new loans ?)
When UK courts decided that LME warehouse receipts were not a transferable title of ownership, bills of lading were used to support borrowing. A legitimate “bill of lading” is proof that a cargo is being shipped, but fake BLs were used to access financing by a number of traders, with most lending banks not equipped to check physical cargoes.
Lacking physical expertise, some western banks stepped back from commodity financing around 2020. As experts in physical metals trading and better placed than banks to handle the risks in financing smaller metals traders, Trafigura announced in the press that they had secured additional bank funding despite the unprecedented price volatility.
4 | ADMISI - The Ghost In The Machine | Q2 Edition 2023
In February 2023, Trafigura announced a fraud by a long standing customer, who had obtained finance on about 1,100 containers of Ni which, on inspection, turned out to be empty.
Covering the position, Trafigura picked up some Ni briquettes from LME warehouse, and discovered that the bags contained stones. This was another blow to the LME nickel market still reeling from the aftermath of a massive short position in March 2022, and delayed the re-start of the 1am LME Select nickel market as the LME told warehouses to audit the stock of nickel briquettes.
If Trafigura could fall victim to this level of fraud, it’s unsurprising that more banks have cut relationships with certain traders, and tightened procedures around commodity financing.
SLOW BOAT TO CHINA...
Copper has many uses in construction, from electrical to plumbing to heating to roofing, but its role as a funding vehicle really came to the fore after China introduced a round of macro controls to counter the rapid rise in house prices in 2009-2010. Bank lending to the housing industry peaked above RMB 2 trillion in 2010 (6.83RMB =$1 in late 2009), falling towards RMB 1.3 trillion in 2011 and 2012, and the industry tried to make up the shortfall.
Facing restrictions on bank lending, developers found that domestic banks would finance inventory at preferential rates, including LME metal on route to Chinese ports. As a US$ denominated asset, this “secured” lending might cost 3-4% compared to 9-10% for domestic lending to the building sector.
Rather than being solely focused on acquiring and hedging building materials, the commodity trading units of some Chinese construction companies grew into significant funding vehicles for their parent companies, using schemes maybe similar to the following example:
• LME warrants were bought and cancelled, and the buyer would show the resulting warehouse receipt to a mainland Chinese bank as proof of ownership of a US$ denominated physical asset heading for China.
•
After a credit process lasting perhaps 2 weeks, the bank would give transit financing at cheap rates. The banks would then return the warehouse receipt (i.e. the “security”) to the buyer along with the funds, and the financing term would reflect the expected delivery time to the Chinese port.
• With the warrant buyer now holding the funds and the warehouse receipt, there was little to stop them from re-warranting the stocks back onto the LME, and then sifting for new warrants before applying for another round of financing.
• The cost to do this would be the LME sift and cost of carry for a couple of weeks until re-warranting plus, perhaps $5/Mt for re- warranting charges minus any contango.
Table 1: Possible example of a transit finance trade
TRANSIT FINANCE DEAL COPPER PRICE MT
VALUE # MONTHS
FUNDING COST AT "SECURED RATE" PLUS LME COSTS (SEE BELOW) TOTAL FUNDING COSTS
EFFECTIVE FUNDING RATE %/YR c.f . FUNDING A FUNDING COST SA
T DOMESTIC RA VED
TE
EXAMPLE: POSSIBLE LME COSTS 2WEEKS FUNDING RENT $/DAY/MT
COMMISSION 1/16% RE-WARRANT $/MT LESS LME CONTANGO LME COSTS Source: ADMISI
$8,000 10,000
$80,000,000 7 3.0%
$1,400,000 $210,000
$1,610,000 3.45%
10.0% $4,666,667 $3,056,667
2.25% $0.50
0.0625% $5.00 $3.00
$70,000 $70,000 $50,000 $50,000 -$30,000 $210,000
FACING RESTRICTIONS ON BANK LENDING, DEVELOPERS FOUND THAT DOMESTIC BANKS WOULD FINANCE INVENTORY AT PREFERENTIAL RATES, INCLUDING LME METAL ON ROUTE TO CHINESE PORTS.
The suspicion remains that some Chinese banks financed many times the tonnage that was actually shipped, and didn’t ask too many questions about collateral, provided the loans were repaid on time (even if the banks were repaid out of new loans?).
Having been willing to lend cheap funds against expected copper shipments without retaining the documentation until delivery, or checking that delivery had been made, it was only another step for some banks to lend against faxed or photocopies of physical documentation.
For less scrupulous traders, there seemed to be few controls to prevent loans from several different banks being secured against the same expected copper shipment, before dropping that material back onto the LME and sifting again. Extending the time to arrival also helped reduce borrowing costs, with savvy traders possibly choosing copper at the back of an LME queue, “booked” to China on the slowest boat available.
Estimates vary as to the amount of loans made per Mt of Cu held in China but, with scant audit of physical deliveries against loans, quick turnover of paperwork with original documents not held, and borrowers likely to repeat the scheme to repay maturing loans, it’s not impossible for this Ponzi- style shadow banking scheme to generate more than 10x leverage.
Diapason and Goldman Sachs called out this financing from 2010 onwards, suggesting a curb on the financing would see prices collapse as stocks left China. Market rumours circulated, with stories of warehouses selling financed stocks or pick-ups being attempted with photocopied documentation, and Chinese authorities clamped down on copper financing by summer 2013. Very quickly, some LME warehouses in Asia started showing significant cancellation and re-warranting of nickel and tin, with just a fraction of the tonnages leaving the warehouse.
The issue wasn’t just limited to mainland Chinese banks, lending cheaply against metal “security” which had been pledged multiple times. More conservative western banks were also financing metal onshore in China, using western warehouse companies in turn reliant on local warehouse operators with less rigorous controls.
The problem came to a head in June 2014, after metal was reported missing in Qingdao. The Chinese authorities locked down access to Qingdao’s Dagang storage, and the official word was that any problem was restricted to Qingdao and Shanghai didn’t have an issue. Several western firms were exposed, with Mercuria and Citigroup contesting $270m of financed metal in the English High Court in late 2014-15, and a number of western banks suffered losses.
In the aftermath, construction companies struggled to get financing with PBOC slashing interest rates 4 times between November 2014 and July 2015. Western traders moved metal out of mainland China where possible, and the LME launched LME Shield in 2015, an electronic audit system also available for non-LME warehouses. As late as 2017, Access World revealed that forged warehouse receipts were circulating in Asia, although no metal had been released against those documents.
Four years after Qingdao, the lid remains firmly on the metal financing issue, and it’s not widely known how widespread the problem was for domestic banks. At best, banks lent funds unsecured at a fraction of the risk price. At worst, banks would have faced capital losses if a housing collapse had forced them to fight over collateral that wasn’t there. In contrast to the 2008 global financial crisis, aired publically and still being felt 10 years later, China’s shadow banking issue has avoided the spotlight.
Rohan Ziegelaar E:
metals.desk@
admisi.com T: +44(0) 20 7716 8081
16 | ADMISI - The Ghost In The Machine | November/December 2018
17 | ADMISI - The Ghost In The Machine | November/December 2018
Page 1 |
Page 2 |
Page 3 |
Page 4 |
Page 5 |
Page 6 |
Page 7 |
Page 8 |
Page 9 |
Page 10 |
Page 11 |
Page 12 |
Page 13 |
Page 14 |
Page 15 |
Page 16 |
Page 17 |
Page 18 |
Page 19 |
Page 20 |
Page 21 |
Page 22 |
Page 23 |
Page 24 |
Page 25 |
Page 26 |
Page 27 |
Page 28 |
Page 29 |
Page 30 |
Page 31 |
Page 32 |
Page 33