Tag Archives: SocGen

Guest Post: Hanging by a Thread – “Very skeptical” judge – former FBI/SEC official eyes London gold and silver fix lawsuits

Written by Allan Flynn, specialist researcher in aspects of gold and silver. 

Five months have lapsed without decision, since London gold and silver benchmark-rigging class action lawsuits received a cool response in a Manhattan court. Transcripts from April hearings show, in the absence of direct evidence, the claims dissected by a “very skeptical” judge, and criticized by defendants for lack of facts suggesting collusion, among other things.

Judge Valerie E. Caproni, former white-collar defense attorney, SEC Regional Director and controversial FBI General Counsel, presided over oral arguments for motions-to-dismiss totaling 9 hours on April 18 and 20.

Its “based entirely on statistics with no other,” the judge said, pouring cold water on plaintiff’s claims of bank collusion in gold benchmark rigging. Defense attorney scoffed as much at the silver hearing. “There is not a single fact… that shows an agreement between my client and the other alleged conspirators to fix the fix.”

Seven banks are being sued in separate gold and silver class action lawsuits currently before the US. District Court, Southern District of New York. The plaintiffs: gold and silver bullion traders, and traders of various associated financial instruments, allege banks conspired in secret closed meetings, to rig the London PM Gold Fix, and Silver Fix benchmarks during the period from 2001 to 2013.

If the motions to dismiss are allowed, the complaints will be thrown out. Plaintiffs on the other hand, are hoping to crack open the door to “discovery,” where they get to access confidential bank communications and records. Turning the tables earlier in April Deutsche Bank surprised by promising to provide such evidence ratting out its former fellow defendants. “No. I cannot consider it at all,” the judge said, stonewalling plaintiffs’ arguments mentioning the move. The German mega-bank settled claims a week prior to the hearing but is yet to hand over records.

Evidence

Alleged proof of downward price manipulation is revealed by averaged charts showing “Anomalous” price drops from 2001 to 2013 in the London Silver Fix and PM Gold Fix. In sympathy, even as the gold price quadrupled from around $300 to $1200 through the period, something counter-intuitive happened. Plaintiffs’ claim during London hours between the AM and PM fixes, the average price declined for each of the 13 years, bar one being flat in 2013. More pronounced effects were seen in silver.

“Those were compelling charts,” the judge responded to a presentation of evidence during the silver hearing. “I mean, seriously, they truly show that something happened.”

av-normalized-gold-prices
Image courtesy of Commodity Exchange, Inc., Gold Futures and Options Trading Litigation, Dkt No. 14-MD-2548 (VEC)

Defendants argue rather than collusion, the price drops show normal “parallel conduct.” Loads of precious metals producers all needing to sell, and a bunch of savvy bankers hoping to buy low. Plaintiffs say the banks have a near-perfect record betting on the fix outcome. Well of course, they trade on “the best information in the world about supply and demand,” the banks’ attorneys retorted.

Findings

Inclusion of a non-fixing bank UBS, in the lawsuit, is described by defendants as “unfair,” and just to “suck in” a Swiss regulator’s findings. If the Court is to be believed, the FINMA report may be all the cases have going for them. Of its 19 pages, the dominating subject is foreign exchange misconduct, with only a few lines about precious metals. Tip-toeing around the topic of collusion, the report describes a process of “cooperation” between UBS and others in precious metals trading. It says the bank shared sensitive client trade information such as “client names”, “stop loss orders,” and “flow information on large or imminent orders” with “third parties.”

Sharing of client trade information by banks, including UBS, in currency trading, the UK Financial Conduct Authority (FCA) reported in 2014, was for the purpose of collusion. Such communication enabled different banks to plan trading strategies, and “attempt to manipulate fix rates and trigger client “stop loss” orders (which are designed to limit the losses a client could face if exposed to adverse currency rate movements).”

Strong similarities exist between the methods and tools used by currency traders to rig benchmarks, and those used by precious metals traders. Referring to the detailed description of UBS offences including collusion in currency benchmarks, the FINMA report said, “the conduct and techniques inadmissible from a regulatory perspective, were also applied at least in part to PM spot trading.”

steps1
Thurgood Marshall United States Courthouse, inset Judge Valerie E. Caproni

Hanging by a Thread

One “misconduct” the report emphasized in precious metals trading was UBS’ manipulation of the silver benchmark in the banks proprietary trading. “A substantial element of the conspicuous conduct in PM trading was the repeated front running (especially in the back book) of silver fix orders of one client.” According to the FCA, front running is a kind of insider trading.

“transaction for a “person’s own benefit, on the basis of and ahead of an order (including an order relating to a bid) which he is to carry out with or for another (in respect of which information concerning the order is inside information), which takes advantage of the anticipated impact of the order on the market or auction clearing price.”

Commenting on the brevity of plaintiffs’ evidence in silver, and the FINMA findings, Judge Caproni pointed out, “your hanging your hat pretty heavily on one line in that report.”

Attorney for defendants found the report favorably vague:

“Your Honor, I apologize. I must say he has made a misstatement three times. The FINMA settlement, Section 3.2.3 talks about UBS conduct with respect to silver. There’s not a not a word or hint about coordination with any other bank. It is between UBS and one of its customers or maybe more than one but no collusion.”

The judge responded, “We have the FINMA report. I knew that wasn’t exactly what it said.”

Confusion about the Swiss findings wasn’t helped by the events of the day. The conference call in German, reported FINMA boss Mark Branson alluding to perhaps more than the report dare, in, “clear attempts to manipulate precious metals benchmarks.” FINMA declined a request to supply a transcript or recording of the conference call, with a spokesperson responding, “We do not publish a transcript, and besides we have nothing to add to the report.”

A couple of things may help explain the regulators haziness, and thus the challenges for sensitive cases as these. Switzerland is a country long associated with gold and banking. In 1970 Zurich was home to the worlds largest gold market. Most of the worlds’ gold is imported to Switzerland for refining. Consider then this impressive feat: An official inquiry is conducted of Switzerland’s largest bank caught up in a scandal involving “precious metals.” Offenses were identified in its Zurich office. The agency reports “serious misconduct” in precious metals trading including sharing of secret client trade orders with “third parties.” The agency head rallies against manipulation of “precious metals benchmarks.” Action against 11 bank personnel, including industry bans of between one and five years, were brought against two managers and four traders for, “serious breaches of regulation in foreign exchange and precious metals trading.” But yet, the word “gold” is absent from the entire report and announcements.

Secondly, FINMA’s 2015 Annual Report describes the sanctions against four now-former UBS traders. It may concern some, that justice is left to financial market regulators when:

“Four further enhancement actions against UBS foreign exchange traders were discontinued in August 2015. Since there were indications that their behavior had contributed to serious breaches of regulatory provisions, FINMA issued reprimands without taking further action against these individuals.”

The Puzzle

If the cases proceed, US commodity futures markets may also come under scrutiny. The Court spotted a not-so-obvious paradox concerning allegations the banks suppress gold prices. “Why would they drive the price down when they are sitting on I don’t know how much bullion? They are driving down the price of their own asset,” Judge Caproni posed.

Plaintiffs claim the banks hold a majority of short gold and silver futures on the US-based Chicago Mercantile Futures Exchange, paper instruments tied to the value of London silver and gold, which increase in value as the bullion prices fall. The banks argue that it’s impossible to tell from mandatory government filings, which banks prosper during the declines, and to what extent. The judge agrees, and defendants are pleased. “I’m very skeptical they have a well-pleaded factual allegation of what the banks’…COMEX position is,” the judge said.

“Mr Feldberg: Then your Honor has said that much better than I ever could.”

Where direct evidence of collusion isn’t available, antitrust law allows the pleading of additional circumstantial evidence that lends plausibility to allegations. Other circumstantial evidence, or “plus-factors,“ listed by gold plaintiffs, namely problematic antitrust facts arising out of the very clubby arrangement of the fix meetings themselves, failed to impress particularly.

“The Court: But weren’t all of your plus factors just the natural – they are just a function of the fix?..I thought every plus factor you pointed to was just that’s the nature of the fix.”

Last Stand

The unconvinced magistrate was likely close to a decision, before four weeks back when the banks had one last throw at dismissal. The court in the silver case was asked to apply a recent ruling where warehouse aluminum price manipulation was deemed not to have impacted end users of aluminum. Any decision on this in silver will likely impact the gold case also.

The question of standing, or which plaintiff’s are close enough to the alleged activity to have suffered injury, was well discussed back in April. For example the Court put to defense: “I’m not saying the two guys at a swap meet from Ohio would be a particularly compelling class representative, but why wouldn’t they have standing?” Plaintiffs seemed to have convinced that the issue could be decided later, if it gets to that. Judge Caproni just wasn’t sure firstly if all the statistics, and facts complained were plausible enough to infer collusion, reminding frustrated gold plaintiffs where the balance lies.

“Unfortunately for you I’m the one who has to make the decision here.

Mr Brocket: Again, with the greatest respect, I am trying to resurrect this here but, look. Every fact alone doesn’t prove collusion.

The Court: I agree.”

inside-shot-of-thurgood-marshall-courthouse
A courtroom in the Thurgood Marshall Courthouse NYC

Decisions regarding motions to dismiss in the London gold and silver benchmark-rigging class actions against banks, initially expected around the end of August, could come any day sooner or later according to someone familiar with the cases.

 

The above article was first published at Allan Flynn’s website here.

Allan Flynn is a specialist researcher in aspects of gold and silver. He is currently investigating for future publication on the same topic and works in property and commercial architecture when he needs to eat. He holds shares in precious metals producers and banks.

The Charade Continues – London Gold and Silver Markets set for even more paper trading

Today the London Metal Exchange (LME) and the World Gold Council (WGC) jointly announced (here and here) the launch next year of standardised gold and silver spot and futures contracts which will trade on the LME’s electronic platform LMESelect, will clear on the LME central clearing platform LME Clear, and that will be settled ‘loco London’. Together these new products will be known as LMEprecious’ and will launch in the first half of 2017.

However, although these contracts are described by the LME as delivery type Physical, settlement of trades on these contracts merely consists of unallocated gold or silver being transferred between LME Clear (LMEC) clearing accounts held at London Precious Metals Clearing Limited (LPMCL) member banks (i.e. paper trading via LPMCL’s AURUM clearing system).

For example, the contract specs for the LME’s planned spot gold trading state that the LME’s proposed settlement procedure is one of:

“Physical settlement two days following termination of trading. Seller transfers unallocated gold to LMEC account at any LPMCL member bank, and buyers receive unallocated gold from LMEC account at any LPMCL member bank

The range of LME contracts for both gold and silver will consist of a trade date + 1 contract (T+1), aptly named TOM, as well as daily futures from T + 2 (equivalent to Spot settlement) out to and including all trade dates to T + 25. Beyond the daily futures, the suite of contracts also includes approximately 36 monthly futures contracts covering each month out to 2 calendar years, and then each March, June, September and December out to 60 calendar months. The LME / WGC press release also mentions plans for options and calendar spread products based on these futures.

precious

As well as trading electronically on LMESelect, these precious metals futures will also be tradeable via telephone market (inter-office market). Trading hours for the daily contract (TOM) will be 1am – 4pm London hours, while trading hours for all other contracts will be 1am – 8pm London hours, thereby also covering both Asian and US trading hours. Detailed contract specs for these gold and silver contracts are viewable on the LME website. The trading lot size for the LME gold contracts will be 100 ozs, which is significantly smaller than the conventional lot size of 5000 -10,000 ozs for gold trading in the London OTC market (and conventional OTC minimum of 1000 ozs of gold). The planned lot size for the LME’s silver contracts is 5000 ozs, again below the conventional lot size of 100,000 – 200,000 ozs for silver trading in the London OTC market (and conventional OTC minimum of 50,000 ozs of silver).

These LME contracts are being pitched as a real alternative to the incumbent over the counter system of gold and silver trading in London which is overseen by the London Bullion Market Association, an association whose most powerful members are the clearing and vaulting banks in London, namely HSBC, JP Morgan, Scotia, and to a lessor extent UBS and Barclays, but increasing ICBC Standard bank as well. But given that the LME’s clearing will sit on top of the LPMCL clearing system and use unallocated transfers, the chance of any real change to the incumbent London gold and silver market is non-existent. Nor will the trading of these LME products give any visibility into the amount of physical gold and silver that is held within the London Market, nor the coverage ratio between ‘unallocated account’ positions and real underlying physical metals.

Five Supporting Banks

This new LME / WGC initiative is being supported by 5 other investment banks and a trading entity called OSTC. These bank backers comprise US banks Goldman Sachs and Morgan Stanley, French banks Natixis and Société Générale, and Chinese controlled bank ICBC Standard Bank. According to a Reuters report about the launch, the World Gold Council had approached 30 firms about backing the launch, so with only 5 banks on board that’s a 16.6% take-up ratio of parties that were approached, and 83.4% who were not interested.

Earlier this year in January, Bloomberg said in a report said that the five interested banks were “ICBC Standard Bank Plc, Citigroup Inc., Morgan Stanley, Goldman Sachs Group Inc. and Societe Generale SA“, so somewhere along the line Citigroup looks to have taken itself off the list of interested parties, while Natixis came on board. The World Gold Council’s discussions about a proposed gold exchange and its discussions with ‘5 banks’ appear to have begun as early as the 4th quarter of 2014 and were flagged up by the Financial Times on 02 April 2015, when the FT stated that:

“The WGC has hired a number of consultants and spent the past six months pitching a business case for banks to consider the alternative trading infrastructure”

“The World Gold Council…and at least five banks are participating in initial discussions”

Notably, this was around the time that LME found out it had not secured the contracts to run either the LBMA Gold Price or LBMA Silver Price auctions. Note, that all 5 of the LME supporting banks, i.e. Goldman, ICBC Standard, Morgan Stanley, SocGen and Natixis, are members of the London Bullion Market Association (LBMA), with Goldman, Morgan Stanley, ICBC Standard and SocGen being LBMA market members, and Natixis being a full member of the LBMA. Goldman, Morgan Stanley, ICBC Standard and SocGen are also direct participants in the LBMA Gold Price auction operated by ICE Benchmark Administration. None of these 5 banks are direct participants in the LBMA Silver Price auction. Notably, none of these banks except for ICBC Standard is a member of the precious metals clearing group LPMCL. ICBC Standard Bank also recently acquired a precious metals vault in London from Barclays and also joined the LBMA’s Physical Committee (see BullionStar recent blog ‘Spotlight on LPMCL: London precious MEtals Clearing Limited‘ for details). Therefore, ICBC Standard seems to have a foot in both camps.

Unallocated Balances, Unsecured Creditors

Given the long build-up to this LME / World Gold Council announcement, and the fact that these LME spot and futures products were supposed to be a genuine alternative to the LBMA bank controlled OTC trading system, the continued use of unallocated settlement and the use of LPMCL accounts by these planned LME contracts underscores that the LME contract do not represent any real change in the London Gold and Silver Markets.

As a reminder, the resulting positions following transfers of unallocated gold and silver through the LME Clear accounts of LPMCL members essentially means the following, in the words of none other than the LBMA:

“Unallocated account basis. This is an account where the customer does not own specific bars, but has a general entitlement to an amount of metal. This is similar to the way that a bank account operates” 

Additional LBMA definitions of unallocated transactions are as follows:

settled by credits or debits to the account while the balance represents the indebtedness between the two parties.

“Credit balances on the account do not entitle the creditor to specific bars of gold or silver or plates or ingots of platinum or palladium but are backed by the general stock of the precious metal dealer with whom the account is held: the client in this scenario is an unsecured creditor.

Alternatively, a negative balance will represent the precious metal indebtedness of the client to the dealer in the case where the client has a precious metal overdraft facility.

Should the client wish to receive actual metal, this is done by “allocating” specific bars, plates or ingots or equivalent precious metal product, the metal content of which is then debited from the unallocated account”.

LME bows to LPMCL

However, it should come as no surprise that these LME spot and futures contracts haven’t taken a new departure away from the entrenched monopoly of the London gold and silver clearing and vaulting systems, for the LME specifically stated in quite a recent submission to the LBMA that it will never rock the boat on LPMCL’s AURUM platform. When the LME presented to the LBMA in October 2014 in a pitch to win the contract for the LBMA Gold Price auction (which it didn’t secure), the pitch said that a centrally cleared solution “would only be introduced with market support and respecting LPMCL settlement“. [See right-hand box in below slide]:

LME potential credit models

In the same pitch, the LME also stated that:

LME Clear fully respects existing loco London delivery mechanism and participants

[See bottom line in below slide]:

LME Pathway to cleared solution

Interestingly, following the announcement from the LME and the World Gold Council, the LBMA provided a very short statement that was quoted in the Financial Times, that said:

“The LBMA saw the announcement with interest and reconfirms it has no direct or indirect involvement in this project”.

While that may be true, what the LBMA statement didn’t concede is that 5 of its member banks, 4 of which are LBMA market makers, do have a direct involvement in the LME / World Gold Council project. Nor did the LBMA statement acknowledge that settlement of the planned LME gold and silver contracts will use the LPMCL infrastructure, nor that the LPMCL is now in specific scope of the LBMA’s remit.

Recall that in October 2015, the LBMA announced that:

“the London Precious Metals Clearing company took part not only [in the LBMA] review, but we have now agreed to formalise our working relationship, with the LBMA providing Executive services going forward. I’m grateful to the LPMCL directors for their leadership and their support for removing fragmentation from the market.”

With the LME contracts planning to use LPMCL, this ‘new dawn’ view of the LME / World Gold Council initiative is in my view mis-guided.

Even COMEX has more Transparency

Anyone familiar with the rudimentary vaulting and delivery procedures for gold and silver deliverable under the COMEX 100 oz gold and 5000 oz silver futures contracts will know that at least that system generates vault facility reports that specify how much eligible gold or silver is being stored in each of the designated New York vaults, the locations of the vaults, and also how much of the eligible gold or silver in storage has warehouse warrants against it (registered positions). The COMEX ‘system’ also generates data on gold and silver deliveries against contracts traded.

However, nothing in the above planned LME contract specs published so far gives any confidence that anyone will be the wiser as to how much gold or silver is in the London vaults backing up the trading of these spot and future contracts, how much gold or silver has been converted post-settlement to allocated positions in the vaults, nor how much gold or silver has been delivered as a consequence of trading in these spot and futures contract, nor importantly, where the actual participating vaults are.

This is because the LMPCL system is totally opaque and there is absolutely zero trade reporting by the LBMA or its member banks as to the volumes of gold and silver trading in the London market, and the volumes of physical metals held versus the volumes of ‘metal’ represented by unallocated account positions. Furthermore, the LBMA’s stated goal of introducing trade reporting looks as dead as a dodo, or at least as frozen as as a dodo on ice.

LBMA stall on Trade Reporting, LPMCL clear as Mud

On 9 October 2015, the LBMA announced that it had launched a Request for Information (RFI) asking financial and technology providers to submit help with formulating solutions to deficiencies which regulators thought the London bullion market such as the need for transparency, and issues such as liquidity that had supposedly been recommended as strategic objectives by consultant EY in its report to the LBMA, a report that incidentally has never been made publicly available. On 25 November 2015, the LBMA then announced that it had received 17 submissions to its RFI from 20 entities spanning “exchange groups, technology firms, brokers and data vendors”.

On 4 February 2016, the LBMA then issued a statement saying that it was launching a Request for Proposals (rRfP) and inviting 5 of these service providers (a short-list) to submit technical solutions that would address requirements such as an LBMA data warehouse and that would support the introduction of services such as trade reporting in the London bullion market. The RfP statement said that the winning service provider would be chosen in Q2 2016, with a planned implementation in H2 2016.

However, no progress was announced by the LBMA about the above RfP during Q2 2016, nor since then. The only coverage of this lack of newsflow came from the Bullion Desk in a 27 May article titled “Frustration Grows over London Gold Market Reform” in which it stated that the 5 solution providers on the short-list were “the LME, CME Group, the Intercontinental Exchange (ICE), Autilla/Cinnobar and Markit/ABS“, and that:

“the pace at which the LMBA is moving forward are causes for consternation in some quarters of the sector”

A quote within the Bullion Desk article seems to sum up the sentiment about the LBMA’s lack of progress in its project:

“It’s not going to happen any time soon. Look at how long it’s been going on already,” another market participant said. “Don’t hold your breath. It seems like we still have a long way to go.” 

What could the hold up be? Surely 17 submissions from 20 entities that were whittled down to a short-list of 5 very sophisticated groups should have given the LBMA plenty of choice for nominating a winning entry. Whatever else this lack of progress suggests, it demonstrates that increased transparency in London gold and silver market trading data is not going to happen anytime soon, if ever.

Furthermore, the opacity of the London clearing statistics that are generated out of the LPMCL clearing system need no introduction to most, but can be read about here.

Conclusion

According to the LBMA, ‘Loco London’ “refers to gold and silver bullion that is physically held in London“, however, given the secrecy which surrounding trading data in the London gold and silver markets, and the lack of publication by any bank about the proportion of unallocated client balances in gold or silver that it maintains versus the physical gold or silver holdings that it maintains, this ‘loco London‘ term appears to have been abused beyond any reasonable definition, and now predominantly refers to debit and credit entries in the virtual accounting systems of London based bullion banks. Nor, in my opinion, will the LME contracts change any of this. One would therefore be forgiven in thinking that the real underlying inventories of gold and silver in the London market and their associated inverted pyramid unallocated account positions are too ‘precious’ to divulge to the market. The Bank of England is undoubtedly licking its chops to the continued opacity of the market.

And its not just my opinion. This latest LME / World Gold Council / investment bank announcement has generated other skeptical reactions. The last word goes to Jim Rickards, who tweeted this in reaction to the latest LME / World Gold Council news:

The pre-2015 London Gold Fixings – More technologically advanced than reported

The financial media has recently pitched the transition of the London daily gold fixings to an ICE Benchmark Administration (IBA) platform as a quantum leap from an antiquated Victorian-era process to a futuristic 21st century electronic auction.

For example, the Wall Street Journal recently said that:

“Four of the banks…had participated in the conference call used to determine the daily fixes, a system largely unchanged for nearly a centuryand that “Gold is the last of the precious metals to make the switch to an electronic platform.”

The evidence suggests however, that in the last decade, the technology utilised in the daily gold fixings was far more advanced than the media commentaries imply, and that since 2004, the old gold fixing was not as technologically backward as is generally accepted.

new court

Rothschild Departs, Barclays Joins – 2004

In April 2004, NM Rothschild announced that it was pulling out of commodity and gold trading, and also stepping down from chairing and participating in the twice daily London Gold Fixings. This left four banks as members of the fixing process, namely HSBC, Deutsche Bank, Scotia Mocatta, and SocGen.

According to Risk.net at the time, “the withdrawal of NM Rothschild from the market forced the London Gold Market Fixing company to introduce new fixing arrangements.

From a practical standpoint, with NM Rothschild no longer part of the fixings after May 2004, the meetings could no longer use Rothschild’s offices in St Swithins Lane near the Bank of England. Another practical point was related to the location of the remaining participants’ offices.

Since Barclays Capital, who took over the fixing seat from Rothschild, was based in Canary Wharf (15-20 minutes train ride east of Bank), the five fixing members were not all located in walking distance of a central physical meeting place in the City of London. Scotia’s and Deutsche’s offices were in the City, but another gold fixing member, HSBC, had also fully moved to Canary Wharf circa 2003. Round trip travel from Canary Wharf to Bank twice a day, or vice versa, would have been prohibitive on all but a temporary basis.

Web-Based Commentary

Rothschild’s departure precipitated discussion of three changes to the Fixings process, specifically, 1) an annually rotating chairperson, 2) a conference call, and 3) a far less well-known, ‘web-based commentary’.

On 29th April 2004, Tim Wood of Mineweb.com wrote an article titled “London Gold Fixing Ritual to End”. The article explained the three changes and referred to the web-based commentary:

“As expected, the London Gold Fixing has announced that it will in future rotate the chairmanship of the arrangement and end a tradition of meeting in person to set bellwether gold prices twice a day.

Starting in May, each member bank will assume the chairmanship of the fixing for a one year period starting with ScotiaBank division ScotiaMocatta.

As of the same date, the Fixing will take place by telephone and the five member firms will no longer meet face-to-face as has previously been the case. As part of this change, it is intended that a web-based commentary of the Fixing will be introduced later this year“, the Fixing said in a statement.

The decision by N.M. Rothschild & Sons to quit the gold business leaves a vacancy at the Fixing. Ongoing members are Deutsche Bank, HSBC, and Société Generale.

Simon Weeks is the chairman-elect of the London Gold Fixing.”

[Coincidentally, Tim Wood, currently executive director of Denver Gold Group, has now ended up sitting on the new 2015 LBMA Gold Price Oversight Committee with Simon Weeks, nearly 11 years after the above article was written.]

On 5th May 2004, the twice daily Gold Fixings transitioned from physically attended meetings at Rothschild’s offices to remote conference calls with Scotia as the new chair.

New York Times: Live web-based commentary

The New York Times, in a 6th May 2004 article titled “Pricing Gold but No Longer Standing on British Tradition” mentions a live” web-based commentary for the daily fixings:

“The London Bullion Market Association, which controls the price-setting process, plans to introduce a live Web-based commentary on the daily price-setting this year.”

‘Nothing was that much different apart from the fact that we didn’t walk down to St. Swithins Lane,’ said Simon Weeks, director of precious metals and foreign exchange at ScotiaMocatta, a unit of the Bank of Nova Scotia.”

(Note: The NY Times meant LGMFL, not LBMA, but they may have got confused because Simon Weeks was chairman of the LBMA at that time, as well as being chairman of the Gold Fixing company LGMFL).

Barclays then completed the purchase of Rothschild’s gold fixing seat in May 2004 (Risk.net 26th May 2004). In its article, Risk.net also confirmed the planned web-based commentary:

“LGMF (London Gold Market Fixing) said it intends to introduce web-based commentary of the fixing later this year.”

Barclays then joined the fixings on 7th June 2004.

Bank of England refers to a Web-based application

The most authoritative confirmation of this “web-based feature commentary” comes from the May 2004 edition of the Bank of England Quarterly Bulletin which was kept in the gold fixings loop as per usual, and saw fit to review and report on the changes taking place in the Gold Fixing. See page 14 of pdf where it states:

“Since 5 May, a telephone conference call has replaced the twice-daily physical meetings. A web-based application to allow viewing of the fixing process is to be introduced later in 2004.

Bank of England Quarterly Bulletin screenshot, May 2004:

BoE web based fixing app
Fast forward to 2014, and a publication titled “Financial Markets and the ACI Dealing Certificate 310-102“, by Philip J L Parker (ISBN 978-1-291-50352-4) also mentions this web-based commentary for the Fixing, stating:

“With effect from May 2004, the traditional face-to-face meetings (previously at the offices of NM Rothschild and Son), were replaced by a telephone fixing procedure. As part of this change, a web-based commentary of the Fixing has been introduced.

(Page 208: London Gold Fixing)

The above publication (published in August 2014) is a training manual for the ACI Dealing Certificate foreign exchange and money markets examinations, which are offered by the wholesale financial market association ACI (now called ACI – The Financial Markets Association). The first edition of this manual was published in 2013.

2014, page 208:

ACI Dealing - gold fixing - web-based commentary

So it appears that a live web-based commentary / web-based application has been used by the five members of the London Gold Fixing since 2004 that allowed the viewing of the fixing process, which would presumably mean viewing the orders entered by each participant and the intra-auction prices. However the existence of such as web-based application is never mentioned by the financial media, who persist in only ever mentioning a conference call, often in conjunction with the words ‘tradition’, ‘antiquated’ or ‘unchanging since 1919′ etc.

Pens and Paper?

It stands to reason that a live web-based application would be introduced and used during a conference call of the daily Gold Fixings. Given that the trader participants were located remotely from each other, it would be essential for the traders at each of the five firms to be able to see prices and current orders on their desktop screens during the fixings, and also essential for final order data to be captured in a trade capture system, then matched, and then sent downstream within trade, clearing and settlement processing systems.

As well as using phones, everything an investment bank trader or an inter-dealer broker does involves using one of their, often, six or more screens as input and output devices. They do not just use ‘bits of paper’ to record orders and trades and then pass these bits of paper to some junior person to run around the precious metals trading desk with. Trading screens are always used in conjunction with phones. Every order has to be captured and displayed, as well as calculated and processed in trade and settlement processing systems and downstream P&L and reporting systems.

We are talking here about order entry, trade execution, trade capture, trade processing, and trade clearing and settlement. We are also talking here about the most sophisticated investment banks on the planet, with the largest and most cutting edge technological and financial resources in any industry. We are talking about HSBC, ScotiaBank, Deutsche Bank, Barclays and Société Générale, not about two-bit bucket shops.

Until August 2014, the daily Silver Fixings comprised three of the same members as the daily Gold Fixings, namely HSBC, ScotiaMocatta and Deutsche. Given that both gold and silver trading would be run from the same precious metals trading desks in these banks,  it seems reasonable to suggest that any technological order capture and display systems that were being used in the gold fixings, would also be used in the silver fixings.

It therefore makes the following claim from Harriet Hunnable of the CME Group hard to fathom when she commented last October on how the CME had taken the Silver Fixings out of the dark ages (CME ‘proud’ of silver fix system):

“In a very short time, we’ve taken a market that was doing this on pen and paper on the telephone to an electronic platform.”

I find this ‘pen and paper’ reference extremely hard to believe given the discussion of a web-based application in the Gold Fixings since 2004. Financial media commentaries at the time, in August 2014, also stuck to the dark ages script with CNBC headlining its coverage as “Victorian-era silver fix joins electronic age“.

Note that even ‘voice-brokered trades’ done by the large inter-dealer brokers such as ICAP and Tullet Prebon make use of screens as well as phones. Screens are intrinsic to all modern voice trading, as are messaging apps, and chat apps (although messaging and chat apps will probably be more highly regulated and  subject to stricter compliance controls going forward).

It would be naive of anyone to think that daily Gold Fixings involving five distinct dealing rooms of five huge investment banks were not using various forms of order entry, trade capture, and various types of networked technology, to keep track of gold and silver fixing prices and orders and to visually display this updated data on traders’ screens and desktops during the daily fixing auctions.

Furthermore, the resulting net order data would have to be passed to other trade processing systems for downstream processing into London Precious Metals Clearing Ltd’s (LPMCL) metal clearing AURUM system, for netting and clearing and settlement, while the price and time-stamp data would need to be passed to price data vendors for distribution as well as to the LGMFL goldfixing.com web site.

Trading floor screens

Without a functional specification document, its hard to know what the original specification of a 2004 web-based commentary/web-based application used on the five trading floors would have entailed, and whether it would be originally designed and built in-house by one or a number of the technology departments of the five fixing member banks, or whether this type of project would have been outsourced. But trading floor technology is always changing and evolving and indeed, trading technology did change rapidly from 2004 to 2014.

Applications designed and used within investment banks do not stay static and they also have to be supported and maintained. Applications either evolve with the evolution of an investment bank’s technology environment or they are decommissioned and replaced. So it’s doubtful if a web-based app created in 2004-05 would still exist in its original version 1.0 form in 2014-15.

Goldfixing.com

Examining the observable technology connected to the London Gold Market Fixing Company also brings up some interesting information. One window into the London Gold Market Fixing Ltd was its website www.goldfixing.com. The domain lookup for the www.goldfixing.com provides both registrant and technical support information.

The site was registered on 22nd December 1999 by  Emilie Rivoire of NM Rothschild (emilie.rivoire@rothschild.co.uk). This would make sense since NM Rothschild was the permanent chair of the daily gold fixings until 2004. The first version of the goldfixing.com website was created by a South African company called Catics Ltd in 2000.

Gold Fixing Rothschild Hackwood

As Catics stated in their scope for the brief of the Rothschild gold fixing website:

“Rothschild, with approval from the other 4 members, approached us to design an elegant new web site. The site was created as a quick up-to-date historic guide about the London Goldfix. All interested parties can see how the price of gold gets fixed twice a day.”

The key requirements for the website included:

  • Provide a graphical view that would indicate the five members buying, holding or selling gold.
  • Build an interactive charting facility so that users can chart historic gold fixes.
  • Integrate site with Rothschild CMS (Content Management System).

Five fixers

Catics Lts also created an updated version of The Rothschild Family website http://www.rothschild.info/ (2003) and also created The Rothschild Archive website http://www.rothschildarchive.org/ (2001, 2003).

Barclays and Sapient

When NM Rothschild departed from the Gold Fixings in 2004 and sold its fixing seat to Barclays, it appears that Rothschild also handed over the responsibility for the website to Barclays, who at some point employed Sapient in a technical capacity for the website. The domain lookup for the site most recently lists a technical support contact for the website of Sapient, with an address of Eden House, 8 Spital Square, London E1 6DU, and an email contact of barclaysmsosupport@sapient.com.

Sapient barclaysmsosupport

Eden House is the London office HQ of Sapient Global Markets. Sapient Global Markets is part of the Publicis.Sapient group, and provides various financial market consultancy and technological services to “capital and commodity market participants” including numerous financial exchanges and clearing houses. Publicis Groupe acquired Sapient in November 2014.

The Sapient phone number +91 1246724778 is an Indian-based number of a Sapient Nitro office in sector 21 of Unitech Infospace in Gurgaon near New Delhi. Sapient Nitro is another division of the Publicis.Sapient Group.

The ‘MSO Support’ in barclaysmsosupport@sapient.com refers to Managed Service Operations (MSO), which is an area within Sapient Nitro’s systems integration practice, which operates from various places including Gurgaon in India. This MSO support team was responsible for the www.goldfixing.com web site that was permanently switched off on the morning of 23rd March 2015.

That Sapient was responsible for the www.goldfixing.com web site is a fact because their indian team in Gurgaon confirmed to me early on the morning of 23rd March that the website had been shut down, as follows:

goldfixing thanks and regards

Furthermore, on their email to me, Sapient (Gurgaon) used the following two Sapient email addresses connected to the Gold Fixing and the goldfixing.com website:

londonpricefixing Sapient

Barclays MSO Support Sapient

A managed service operations team would generally be responsible for content management and delivery, as well as underlying web applications and servers etc.

Before the plug was pulled on the www.goldfixing.com website, fixing prices and associated trading data and gold bar quantities always appeared rapidly on the GoldFixing website straight after the 10.30am and 3.00pm fixings were completed, along with accompanying timestamps down to the exact second.

For example, on 23 October 2014, the morning gold fixing completed at 10:31:16. This information was rapidly updated on to the goldfixing website, as well as being sent out to all the major data vendors such as Bloomberg and Thomson Reuters:

goldfixing snapshot daily price and timestamp

Distribution of near real-time price data could not have been done without an electronic system that captured the fixing data, stored it in a database table, and fed it to a front-end website query.
Likewise, the historical price, bid-offer, bar total and date data which were viewable on the goldfixing website would also have needed to be stored in a database table and accessible via a website query. For example, see the last historical data of gold fixings  from 18th and 19th March 2015 to be displayed on the old goldfixing website before it was switched off:
gold fixing website historial previous days data

This fixing data that appeared on the goldfixing website has to have been supplied by other connected systems such as a fixings order capture and processing system. There cannot be website outputs within inputs, which by definition implies that there are also calculations performed as well as storage and retrieval. i.e. information systems and not ‘pencils’ and ‘bits of paper’ as some of the financial media seem to think the modern daily fixings made use of.

Managed Service Operations (MSO) offerings from companies such as Sapient, often include software/services that facilitate collaboration, and there are also lots of ready-made collaboration applications available on the market. For example, Microsoft Online Services is a server hosted enterprise software suite that can include Office Communications Online, Microsoft Office Live Meeting, and Sharepoint Online. Suffice to say, these products/services (which can be locally or cloud hosted) provide on-line real-time instant messaging and communications (Microsoft Communications Online), live conferencing with video and audio and messaging (Microsoft Live Meeting), or a collaboration platform (Microsoft Sharepoint Online). Citrix also offers a lot of products/solutions in this space such as GoToMeeting.

So some of the above types of software/services would fit the bill for providing precious metals traders’ workstations with web-based commentary, and messaging and communication apps that could be used in the daily fixings alongside phones. Outputs from some of the above could also be integrated into web site price data feeds through messaging middleware.

But there is another more important connection between the London Gold Market Fixing Company and Sapient Global Markets which points to another Sapient app being more than a web-based ‘commentary’.

 

The replacement Gold Fix – Request for Proposals

When the London Gold Market Fixing Limited (LGMFL) and the London Bullion Market Association (LBMA) launched a Request for Proposals (RfP) to administer the new LBMA Gold Price auction on 4th September 2014, Sapient Global Markets was one of the applicants to submit a proposal. This proposal was submitted in conjunction with Autilla Ltd. Previously, for the silver fixing replacement in mid 2014, Autilla initially submitted a standalone proposal, and then in the final week in early July, teamed up with the London Metal Exchange (LME) on a joint bid. Interestingly though, for the gold fixing proposal, Autilla joined up with Sapient in a joint bid on Day 1, so Autilla must have deemed a joint bid with Sapient as being advantageous.

Out of eight proposals received, the Autilla/Sapient proposal was among five proposals to get short-listed by the LBMA, and although they didn’t win the new contract, Autilla/Sapient did make a presentation of their proposal at the LBMA closed-door ‘market’ seminar on 24th October which saw presentations by the five short-listed parties. Note also that there was a third member of this Sapient/Autilla partnership called ‘Global Rate Set Service’, which was also referred to in the proposal as ‘Global Rate Set System’ and ‘GRSS’. This appears to be Global Rate Set Systems, a New Zealand based company.

Sapient letterhead

In the Sapient/Autilla proposal summary,  which takes the form of  a 2-3 page letter to the LBMA dated 27th October, Page 2 describes a ‘current process‘ and also modifications for the new proposed process.

Reading Page 2 of the proposal, its clear that Sapient are intimately familiar with the ‘current process‘ and they only suggest ‘making changes’ to the current process where needed. Sapient state:

“Our solution is one that has a look and feel which is easily recognisable and known to those already familiar with the current process.”

Sapient’s reference to an ‘easily recognisable and known‘ ‘look and feel‘ of its proposed system suggests that ‘those already familiar with the current process‘ were familiar with a similar system.

‘Look and feel’ is a term that’s most commonly used in software development and nowadays rarely means anything outside the software industry. Just google ‘look and feel’ with or without the quotes to see what I mean. In software solutions, ‘look and feel’ will almost always mean “the appearance and function of a program’s user interface”, or “the design and formatting of a graphical user interface (GUI).”

Sapient is saying that those who were using the current process at that time in October 2014 (i.e. the traders of the remaining four fixing members ) would recognise and know an existing graphical user interface that they were familiar with when looking at Sapient’s proposed new graphical user interface.

Sapient states that it has ‘kept’ seven ‘main functions’ of the current process, and then goes on to list the functions that it has kept; these functions include participants logging in, participants entering indicative bar Buy, Sell and No Interest orders in bar amounts, a virtual Flag, matching within tolerance (50 bars), sharing out bars within tolerance, and fx rate pricing:

Sapient app functionality

Sapient then lists the “new or modernised‘ ‘changes’ it is proposing ‘to achieve additional objectives of modernisation, transparency and regulatory cover‘. These new or modernised changes include house and client trades, intra-round price determination, real-time pricing commentary for full distribution, and a GUI messaging portal. Connecting in to the fixing via the messaging portal suggests that any previous messaging would have been done through standalone messaging/chat apps (like those used by interest rate and fx traders).

Interestingly, the Sapient proposal refers to automating some of the tasks that were done by the chairman of the Fixings which sounds like this entailed releasing the final fixing orders for matching, and then processing trade confirms etc.

Sapient new functions

In its proposal to the LBMA, Sapient therefore appears to be describing an existing electronic networked order capture and processing system that the gold fixing process was already using (up until Thursday 19th March 2015). It makes perfect sense then that Sapient had the contract to run the www.goldfixing.com website if it was also responsible for building, maintaining and supporting other parts of the recent gold fixing technical architecture.
Gold Fixing Document Retention Policy

That networked technology was used within the daily gold fixings prior to the transition to ICE’s WebICE is also supported by the  requirements of the “Document Retention Policy” of the London Gold Market Fixing Company, dated 29 October 2014.

This Document Retention Policy, in section 2.3, states that the chairperson of the fixing process is responsible for keeping a record of the following data: member firms participating on each call, names of the individuals from each firm, opening price and sources of opening price, prices tried during the fixing, “bid and offer figures of each member firm at each price tried”, final fix price in dollars, euros and pounds, the time the price was fixed, euro and sterling exchange rates used to determine the fix price, and volume of transactions executed between participating member firms. That is a lot of data to have to record manually twice, each and every day, so again, this suggests that the chairman was not recording this information manually.

Note: As part of the application process to run the new gold fixing, all parties who submitted bids to the LBMA, including Autilla/Sapient, had to sign a non-disclosure agreement (NDA) with the LBMA, so it would be difficult to verify the technical details behind the Sapient/Autilla proposal, as they are most likely covered under the NDA and could not be revealed without the permission of the LBMA.

2004: A Price Oddity  –  Web-based App & Gold Price Manipulation?

There is also an interesting correlation between the timing of a web-based application being launched for the Gold Fixing in 2004 and the timing of alleged gold price manipulation beginning in 2004.

On 28 February 2014, Bloomberg’s Liam Vaughan wrote an article titled “Gold Fix Study Shows Signs of Decade of Bank Manipulation“, which stated that:

The London gold fix, the benchmark used by miners, jewelers and central banks to value the metal, may have been manipulated for a decade by the banks setting it, researchers say.

Abrantes-Metz and Metz screened intraday trading in the spot gold market from 2001 to 2013 for sudden, unexplained moves that may indicate illegal behavior. From 2004, they observed frequent spikes in spot gold prices during the afternoon call. The moves weren’t replicated during the morning call and hadn’t happened before 2004, they found.

Large price moves during the afternoon call were also overwhelmingly in the same direction: down. On days when the authors identified large price moves during the fix, they were downwards at least two-thirds of the time in six different years between 2004 and 2013. In 2010, large moves during the fix were negative 92 percent of the time, the authors found.

There’s no obvious explanation as to why the patterns began in 2004, why they were more prevalent in the afternoon fixing, and why price moves tended to be downwards, Abrantes-Metz said in a telephone interview this week.

 Could the introduction of a trading desk web-based application into the fixings in 2004, which would have provided gold trading desks with extra eyes into the auction proceedings, have presented a means for facilitating a type of gold price manipulation which previously was not possible during the purely phone based meetings held at Rothschilds in St Swithins Lane?

 

fca stairs

The FCA, Barclays and Daniel Plunkett

On 23rd May 2014, the UK’s Financial Conduct Authority (FCA) announced that they were fining Daniel Plunkett, a former Barclays trader and director of its Precious Metals Desk, for manipulation of the gold price during the afternoon gold fix on 28 June 2012, and also fining Barclays for breaches of two Principles of the Authority’s Principles for Businesses between 7 June 2004 and 21 March 2013.

The FCA’s ‘Final Notice’ explaining the fining and prohibition of Daniel Plunkett, provided details of Plunkett’s trading into the gold fix on the afternoon of 28 June 2012. The ease and speed with which Plunkett, on two occasions, rapidly placed and then cancelled proprietary trades into the gold fixing during the fixing that afternoon, suggests that he was using an automated order entry system to place and cancels those trades, and also to unwind the second trade after the fixing completed.

Indeed, at that time in 2012, Barclays’ systems did not differentiate between a Gold Fixing trade executed by a Barclays trader and a gold spot market trade executed by that same trader. And since proprietary gold spot trades would be entered electronically, so too would Gold Fixing trades.

According to section 4.14 of the Final Notice document on Plunkett:

At 3:06 p.m., shortly after the Chairman had increased the proposed price to USD1,558.50, Mr Plunkett, who had not placed any previous orders during the Gold Fixing, placed a large sell order of between 40,000 oz. (100 bars) and 60,000 oz. (150 bars), with Barclays’ representative on the Gold Fixing. This order was incorporated by Barclays’ representative into Barclays’ net position, which led to Barclays declaring itself to be a seller of 52,000 oz. (130 bars).”

At 3:07 p.m. Mr Plunkett withdrew his entire sell order, which resulted in Barclays’ representative withdrawing Barclays’ position (selling 130 bars). This reduced the imbalance in the 28 June 2012 Gold Fixing from 190 bars to 60 bars (155 bars buying/215 bars selling)” Section 4.17

At 3:09 p.m., Mr Plunkett again placed a large sell order, 60,000 oz. (150 bars), with Barclays’ representative, who, also taking into account changes in customers’ orders, declared Barclays’ net position in the 28 June 2012 Gold Fixing to be selling 40,000 oz. (100 bars).” Section 4.21

Shortly after the conclusion of the 28 June 2012 Gold Fixing, Mr Plunkett repurchased 60,000 oz. (150 bars) of gold by executing an internal trade with Barclays’ Gold Spot Book. The purpose of executing this order was to unwind the 60,000 oz. (150 bars) position he had taken during the 28 June 2012 Gold Fixing.” Section 4.24

If an internal trade that Plunkett executed with the gold Spot Book could unwind an outstanding trade that he placed into the Gold Fixing, then the two trades, and the manner in which they were input would need to be similar, which, we will see below that they were.

Barclays –  Gold Fixing trades were identical to Gold Spot trades

The FCA also issued a ‘Final Notice’ detailing the background to the financial penalty imposed on Barclays,  for Barclays’ failure to , amongst other things, create systems on its precious metals desk “that allowed for adequate monitoring of traders’ activity in connection with the Gold Fixing”. The Barclays “Precious Metals Desk” was Barclays trading desk responsible for gold, silver, platinum, palladium and rhodium.

“The systems and reports did not formally record orders placed by traders in the Gold Fixing until 5 February 2013 and did not identify Gold Fixing transactions separately from general gold spot trades until 21 March 2013. As a result, Barclays was unable to adequately monitor what trades its traders were executing in the Gold Fixing or whether those traders may have been placing orders to affect inappropriately the price of gold in the Gold Fixing.” Section 2.3

Barclays relied upon systems and reports that did not differentiate between Gold Fixing and gold spot market trades executed by its traders. (Barclays addressed this on 21 March 2013, when it updated its systems to specifically record Gold Fixing trades as such.) This meant that during the Relevant Period, Barclays could not adequately monitor its traders’ orders and trades executed in the Gold Fixing.” Section 4.36

So, section 2.3 and section 4.36 of this FCA Final Notice tells us that in 2012, gold fix trades executed by Barclays traders were seen as identical to gold spot trades executed by those same traders, and that both sets of trades used the same systems. Plunkett was not being monitored and was independently executing trades that were identical to gold spot trades, and these trades were flowing into Barclay’s net gold fixing position. This would have required an electronic trading platform. If Barclay’s house and customer gold fixing trades were on a technological platform in 2012, then the whole notion of the gold fixing orders with the other fixing participants also not being integrated into an electronic platform prior to 2015 is implausible.

The rapidity with which Plunkett engaged actively in the afternoon gold fixing on 28 June 2012 was also reiterated in the FCA’s Final Notice for Barclays:

“On 28 June 2012, a Barclays trader, Mr Daniel Plunkett, participated actively in the Gold Fixing” Section 2.6

“In particular, he placed a large sell order of between 40,000 oz. (100 bars) and 60,000 oz. (150 bars) with Barclays’ representative on the Gold Fixing, then withdrew it completely one minute later and subsequently placed another large sell order of between 40,000 oz. (100 bars) and 60,000 oz. (150 bars) two minutes after that.” Section 2.9

The move by Barclays on 21 March 2013 to finally differentiate between prop trader executed gold spot trades and prop trader executed gold fixing trades, also suggests that whatever the change was, it took an existing transaction type of gold spot trade and reflagged it as a gold fixing trade. These changes would have all been conducted on a pre-existing electronic platform (since the gold spot book was on an electronic platform), again undermining the notion of a purely pens and paper supported approach to the gold fixing using a system largely unchanged for nearly a century.

Interestingly, neither of the FCA Final Notices issued in connection with Barclays, Plunkett and the gold price, nor any other FCA comments on its investigation into precious metals manipulation in London, make any reference whatsoever to whether the FCA examined  precious metals traders’ messaging app logs or other trader online communication dialogues. This is odd given that messaging apps were seen to have been widely used by all other traders in the recent LIBOR and FX price manipulation scandals. See here for some LIBOR examples and here for some FX examples of trader transcript manipulation chats.

Given that there appears to have been a web-based commentary in the gold fixings since 2004, as well as very sophisticated gold fixing order and price data capture in Barclays systems and in the most recent iteration of the Sapient supported goldfixing website, perhaps the financial media can take a look into this before claiming with certainly that the gold fixings only went on to an electronic platform during the 20th March 2015 transition to the ICE/IBA/LBMA Gold Price architecture.