LBMA Data: Beyond The Smoke And Mirrors
Today’s AM fix was USD 1,285.75, EUR 972.34 and GBP 847.84 per ounce.
Yesterday’s AM fix was USD 1,323.75, EUR 999.20 and GBP 870.29 per ounce.
Gold fell $12.50 or 0.94% yesterday and closed at $1,310.30/oz. Silver also followed suit and dropped $0.16 or 0.81% and closed at $19.66.
Gold rose in early Asian trading, then fell back to lose some of its early gains on the back of better than expected economic data. U.S. manufacturers reported best figures since June 2011. Silver finished the day with a gain of 0.81%.
LBMA Data: Beyond The Smoke And Mirrors
On Wednesday, the London Bullion Market Association (LBMA) published June clearing turnover statistics for the over-the-counter London Gold Market which revealed a new multi-year record of astronomical monthly ‘gold clearing’ activity for June and rounded off a quarter of record activity from April to June, a period during which the gold price fell substantially. These monthly statistics are usually released at least one month after the end of the reporting month, hence the latest figures were released on July 31st.
On a six month comparison from January to June 2013, even though the gold price fell by 20.7% over the period, demand for gold shot up by 39%. As the LBMA commented on Wednesday:
“demand for gold has increased by 39% since January, buoyed by strong physical demand particularly from China and India which has more than offset sales by ETF funds in the western economies“
June gold clearing figures broke a new record for millions of ounces transferred with on average 29 million ounces transferred (cleared) per day in the London Market, improving on the previous multi-year record of 28.2 million ounces transferred per day which was reached in May.
Since the gold price trended lower in June compared in May, the average daily value in US dollar cleared was lower in June than in May, even though volume was higher. The June average daily number of gold transfers reached 4,624, which although less than the daily average of over 5,000 in both May and April, is still one of the highest on record.
Average daily ounces transferred in the second quarter at 27.1 million were 22% higher than the first quarter daily average of 21.1 million ounces. Likewise, the average daily number of transfers in the second quarter at 5,035 was 61% higher than the average for the first quarter of 3,121. Therefore there has been a significant inverse correlation between the gold price and clearing (as well as trading activity) in the second quarter. As the gold price dropped clearing activity intensified. This does not prove causation but it’s significant nonetheless.
It’s constantly a source of amazement to the gold world that these LBMA average daily clearing statistics are so large. For example, 29 million troy ounces is equivalent to 902 metric tonnes of gold being cleared each day through the London Market. Even more startling is that these figures are only for clearing and not for trading.
In 2011 the LBMA conducted a survey of trading activity amongst its gold trading members to gauge the liquidity of the London Gold Market. The survey found that there was a roughly 10 to 1 ratio of trading volume to clearing volume and that the average daily trading volume for the London market during Q1 2011 was 173 million ounces of gold. This was for a three month period (January to March 2011) during which the average daily clearing volume was 18.7 million ounces. This large 10 to 1 multiple exists since a lot of trading activity is netted before being cleared, as happens in any financial market.
Therefore with the June 2013 average daily clearing figure of 29 million ounces, using the 10-1 multiplier, the average daily June gold trading volume is in the region of 290 million ounces. This is per day, and equates to 9,020 metric tonnes of gold per day being executed through the London market.
According to the World Gold Council, annual gold mining production was only 2,800 tonnes in 2012 and recycled gold supply only added another 1,600 tonnes, creating annual supply of 4,400 tonnes in total. Given physical supply, then the 9000+ tonnes traded per day is extremely hard to fathom. Therefore, it’s worth examining what these LBMA clearing figures actually represent.
The London Precious Metals Clearing Company
The LBMA gold clearing data represents the volume of gold cleared by the six investment bank members of clearing company London Precious Metals Clearing Limited (LPMCL) through their electronic metals clearing hub called AURUM.
The six members of LPMCL comprise London based entities of HSBC, JP Morgan, UBS, Barclays, Deutsche Bank and ScotiaMocatta, four of which, HSBC, JP Morgan, Barclays and Deutsche, operate precious metals vaults in London (UBS and Scotia having access to vaulting space), and four of which, HSBC, Barclays, Deutsche and Scotia are members of the London Gold Market Fixing Ltd which runs the twice daily London gold fixing meetings.
All six members of LPMCL are also market-making members of the LBMA for gold spot and forward transactions, and all six members have gold account facilities at the Bank of England since they execute bullion trades on behalf of central banks and also engage in gold liquidity management with the Bank of England.
Of the six clearers JP Morgan has a 45% share of the London clearing market (for gold and silver), and can put through up to 5,000 transactions per day although this is during busy days. The other dominant clearing member is HSBC which has been known in the past to represent over half of the gold transactions cleared in the London market. Therefore JP Morgan and HSBC together represent the lion’s share of the London precious metals clearing market.
Monthly Gold Clearing Data Metrics
The monthly gold clearing data released by the LBMA come in three metrics, average daily clearing volume in millions of ounces, average daily clearing volume valued in US dollars, and average daily number of clearing transfers.
The clearing data comprises three components.
1. Firstly, the figures include clearing representing book transfer trades between parties within one clearing members’ book, or book transfers between parties across two clearing members. These book transfers are done through LBMA unallocated accounts where no physical transfer of metal occurs.Book transfers are done loco London, in other words they are for gold trades that settle in the London Market for London Good Delivery gold, but they could be for trades done between LBMA market-makers or trades done by any third parties anywhere in the world and would include location swaps. These other parties could be, for example, mining companies, refiners, mints and jewellery manufacturers.
2. The clearing data also includes physical transfers and shipments of gold by the six clearing members, for example shipping consignment stocks to a jewellery manufacturer in Turkey or bullion purchased in London by the State Oil Fund of Azerbaijan back to Baku.
3. Lastly, the clearing data includes transfers between the six clearing members’ gold accounts maintained at the Bank of England. What this really means is that the six clearers will have net positions each evening between themselves at the Bank of England, and the Bank acts as central clearer for the six clearers, passing book entries between their accounts to reflect the overall net position. This is essentially liquidity management for the six clearers and is rather opaque since the Bank can add supply, operating outside normal market forces.
Although the LBMA does not release a percentage breakdown of the three clearing categories that comprise the monthly clearing statistics, it’s possible to infer the first category percentage since book entries are through unallocated accounts.
In April 2013 the LBMA in conjunction with the London Platinum and Palladium Market (LPPM) and HM Revenue and Customs (HMRC) drafted a VAT treatment document for the London precious metals markets. It revealed that in the London precious metals market “in 95% of trades, trading in unallocated metals” was undertaken, and because unallocated metal transactions are treated as services for VAT purposes, “the location of the underlying metal is not relevant”.
If 95% of gold trades in the LBMA are for unallocated metal then the clearing statistics should also reflect this and at least 95% of the clearing statistics would also be for non-physical or book entry movements. This would mean book entries represent at least 855 tonnes per day of the daily 900 tonne clearing figure, and 8550 tonnes of the daily trading figure. It’s unclear as to the size of the transfers between the clearers’ accounts at the Bank of England but these are also undertaken through the Bank’s unallocated accounts for the clearers.
If the average trade size for an unallocated trade is larger than the average trade size for an allocated trade, then the total percentage of unallocated metal attributed to the London Market would be a lot higher than 95%, based on a simple weighted average calculation. This would be a realistic assumption given that, at 290 million ounces of gold traded each day in June 2013, 5% allocated gold daily trading would comprise 14.5 million ounces or 451 tonnes per day of allocated gold each trading day, which is extremely high.
There are only about 720,000 good delivery bars (9,000 tonnes) stored in the London Market, the majority of which is vaulted in the Bank of England’s 10 bullion vaults, and to a lesser extent with the HSBC custodied gold ETF (900 tonnes) and at other commercial storage facilities. With only 4500 tonnes of global gold supply each year, a maximum daily average of 20 tonnes of ‘new’ supply could hit the London Market each trading day, and not all gold passes through London, with a lot in other centres active in physical gold such as Zurich, Singapore, Hong Kong, Dubai and New York. Therefore a more realistic 10 tonnes of new supply gold could enter the London Market each day. Added to this would be normal physical gold trade flows from participants buying and selling and shipping the gold to and from London, but remember that the London clearing statistics do not include physical movement of gold arranged by the clearers at locations other than London.
What Gets Weighed Gets Recorded
In a recent LBMA assaying and refining seminar in March of this year, the transition and testing from a beam balance scales to an electronic scales for the London Gold Market was discussed. This seminar highlighted that about 75,000 bars (937 tonnes) were tested on the new scales by The Bank of England and the commercial vaults over the 2010-2012 period. This represented some of the gold either arriving into or departing London. The Bank of England also said that it had “weighed approximately 400 tonnes over the last year” on the new scales, which would presumably be gold on the move as opposed to gold held in long term storage. Although the new scales have not yet been fully introduced, these tonnages quoted give a more accurate representation of physical gold flowing in and out of the Bank and in and out of the London market.
Therefore, in the clearing statistics, the category occupied by “physical transfers and shipments of gold by the six clearing members”, patently cannot be in the order more than 1% of the trading statistics and by inference, cannot be more than 1% of the clearing statistics.
The LBMA clearing statistics therefore essentially represent huge daily trading through unallocated accounts, most of which is classified as spot delivery, but which is backed by very small physical metal foundations. The clearing statistics while interesting, need to be made more transparent and granular beyond the headline data. Otherwise they tend to obscure rather than illuminate.