Clive Furness has seen the future of open outcry, and it looks a lot like the present of electronic trading. A former director of business development at the London Commodity Exchange and then the London International Financial Futures Exchange, Furness now runs Contango Markets Ltd, an independent consulting firm specializing in derivatives, clearing, technology and e-commerce
“Look at electricity, or freight futures, or carbon emissions,” he says. “Look at any of these markets that came into existence after the advent of screen trading. They all trade without floors, and the ones that meet a hedging need don’t seem to need them.”
And not only do they trade electronically, they even move electronically. Or at least their movement is planned and tracked electronically, thanks to the now mature field of “supply chain management.” In the United States, electronic warehouse receipts have been legal tender since 2000, and the New York Board of Trade (Nybot) has been offering a centralized database and clearinghouse for digital warehouse receipts since 2003.
More intriguingly, a company called Triple Point Technology markets a suite of trade-matching and money management products, among which is a module called the “Visual Cockpit,” which makes it possible for physicals traders to arrange delivery of their product by playing connect-the-dots on their computer screens. Only, the dots are transport hubs and the connections are rail routes, pipelines or maybe trucking companies, depending on whether the product is “wet” and can flow through a pipe or “dry” and needs to be loaded into trucks or trains. Users map out the desired route and the computer tells them if it can be arranged and what it will cost by analyzing the capacities, contracts and transportation costs — all the variables that make logistics so much fun. The system even makes the travel arrangements.
Such automation is also making inroads in managing the “financial supply chain,” harnessing the flow of documentation that accompanies commodities as they pass from hand to hand and mutate from form to form in such a way that commodity companies — and increasingly, financial institutions — trade around their physical positions using asset-backed derivatives. A slew of projects in the software development pipeline promise to streamline the logistics process even further. A recent United Nations report on global commodity flows showed that just-in-time settlement has a long way to go, even though just-in-time delivery had long ago become a reality.
Two widely hyped technologies, Enterprise Resource Planning (ERP) and Electronic Data Interchange (EDI), have just begun to yield benefits. ERP integrates the requirements of all a company’s business departments into one computer system, while EDI provides for the transfer of data between different companies using agreed upon standards. Unfortunately, ERP only adds value within companies and EDI only adds value within jurisdictions. That is changing, as we will see shortly, but even hard-core technology proponents say there are limits.
Greg Leck, Triple Point’s vice president of business development, says cash market deals won’t become completely automated until deep changes take root, either in technology or in human nature. “This is still a business of distrust more than trust,” he says. “There are just too many variables in cash market commodities transactions — too many grades of quality, too many ways to rip someone off.” For now, he says, all the back-office procedures are becoming automated. “Everything to do with moving the deal along once it is made is being automated,” he says. “And everything to do with defining the deal before it is made is becoming automated, but the bulk of the actual deal-making is done by phone. Even when it is then executed on a platform, it will continue to be a person-to-person activity.”
That argument certainly echoes that made by back-month floor brokers, who say the calendar spreads and swap spreads and other intramarket and intermarket trading services they offer hedgers looking to juggle risks simply cannot be replicated by a computer. But the echo is faint to say the least. A commodity futures trade is not really a deal; it is a transaction embedded in a deal. All the subtle negotiations that go into true deal-making have been pre-fabricated by the exchange, which determines quantity, quality, delivery point and the headaches that make cash trading an art rather than a craft.
And commodity futures markets exist to support the cash market. They are, in a sense, a back-office function of the real market — albeit a back-office function that contributes to price discovery. “These floor brokers see themselves [as keeping] lots of abstract relationships in their heads,” says Philip Haynes, head of technology for Computershare Markets Technology, the global share registry company that has leveraged itself into a global provider of trading technology. “They think that is a skill, but in fact it is a function and functions can be replicated.”
Spreads as Securities
“If you are dealing with something like the crack spread, where you are spreading crude oil futures against gasoline futures and heating oil futures, you can basically say that the purchase of a certain number of crude oil contracts implies the sale of a certain number of heating oil and gasoline contracts —and visa-versa,” he says. “It is relatively simple to take a very active spread like that and treat it like a security in itself. This is what floor brokers do when they quote a spread, but with computerized trade-matching you can actually turn the spread into a security and offer it atomically, as a unit, so that it trades in its own space.” As people bid the securitized spread up or down, he says, trades are implied into other markets, and as other markets move, trades are implied into the spread.
“The result is true, hittable bid-offer prices all the time in a wide range of spreads,” he says. “Floor brokers simply cannot offer that — try as they might.” And, he says, implied-trade algorithms can be used in markets that have only a tangential relationship with each other.
In Europe, such algorithms have already led to robust spreading among regional electricity providers and between electricity futures and carbon emission futures. In fact, according to European Energy Exchange Chairman Hans-Bernd Menzel, spreading between regional electricity contracts and the pan-European carbon emissions has been a primary driver of increased correlations among electricity markets far apart from each other.
Obviously, the skill will come in determining the algorithms — or is algorithmic design also just a function? If Haynes is right, retail traders will soon see hittable spreads in the relationship between disparate commodities such as corn and freight rates or soybeans and methane — not to mention between commodities and commodity indexes. He says skilled algorithmic traders will find and exploit correlations on correlations, spreads on spreads, resulting in chains of orders being implied from market to market. “The chains become huge and complex and no floor trader can do this in his head,” he says. “Exchanges will find that they can capture even more volume by creating a whole new class of tradable products built around their core products. You’ll end up with a whole range of found trades as different participants begin finding they are trading the same six-chain implieds.”
Who needs clearing houses?
Computers may not be able to inspect the quality of a batch of corn, but companies like Caliber Inspection and SGS are certainly doing their best to make it easier to carry out such procedures. SGS dates back to 1878, when it launched in Rouen, France, as a grain shipment inspection house. Today, it offers global inspection and benchmarking of commodities as diverse as grains and “ambient air” — the stuff your employees breathe, which makes them slow and sluggish if improperly filtered.
By generating reliable benchmarks for quality and then quantifying quality values for suppliers around the world, these companies are helping to streamline the deal-making process. Another new development is so-called “credit-contingent matching,” which essentially automates mutual credit checks. Such procedures make it possible for two companies dealing in physical commodities to run mutual credit checks in real time. Each company’s credit rating becomes another variable evaluated in seconds as part of a trade execution.
Peter Jessup runs Computershare’s markets technology division and says such technology can offer a viable alternative to using clearinghouses for OTC trades, although companies with unproven credit may find themselves frozen out of such systems. “The real short-term boom in this technology may come in emerging markets, where they don’t have capital to set up a central counterparty,” he says. “In fact, we’ve delivered an automated fixed income solution in the Philippines using a bilateral credit limit model that allows auto-execution of orders contingent on credit availability between the two counterparties, but on an anonymous basis.”
He sees OTC and exchange-traded markets continuing to converge, and predicts a surge in trading among different asset classes. It’s the kind of projection that looks especially promising for energy markets that manage to become interwoven into the electronic infrastructure — at least in light of Menzel’s observation on the correlation among European energy and emissions markets.
Duel in Dubai
The New York Mercantile Exchange (Nymex) aims to become the global trading place for oil of all kinds and from all regions. Towards that end, it’s opening trading platforms in London and Dubai, with more possible. But each platform is an old-fashioned trading pit fighting the global trend. Nymex President Jim Newsome and its London boss Roy Leighton rely on the standard response, “It’s what the customers want.”
We will see, for the Dubai Mercantile Exchange is opening in tangential competition with India’s Multi-Commodity Exchange (MCX), which is also a partner in a joint venture that will offer electronically traded futures in metals and other products. What’s more, the MCX project is a java-based venture, using technology generally trusted for front-ends to construct trade-matching and execution software.
Exchanges and the Real World
Indian exchanges, with double-digit growth and surging retail communities, have certainly generated more than a passing blip on the global radar. Their software programs are certainly innovative, but it is far from clear they are robust, as we will address shortly. And old errors seem to be popping up in the most unfortunate of places. Launching afresh, for example, theoretically allows exchanges in the developing world to skip the whole legacy problem facing U.S. exchanges. That is true technologically, but many Indian exchanges have delivery points for agricultural commodities in the middle of urban areas for no other reason than that it how it has always been done.
The tragedy of such an approach is that a commodity exchange’s success often comes from the degree to which the exchange’s markets are embedded in and serve the underlying cash market. Germany’s Deutsche Terminboerse (today’s Eurex) managed to famously steal the Bund contract away from Liffe back in 1997 by pitting its screen-based platform against London’s pits and slashing fees, but Nymex’s gold contract is holding its own against an electronically traded version at the Chicago Board of Trade, despite that the Chicago product costs less than half as much to execute.
And the quirky London Metal Exchange (LME), with its rotating ring open outcry trading and its myriad short term contracts, is going strong after all these years. One reason, at least in the case of the LME, is it has the most extensive network of warehouses of any metals exchange on the planet and the stocks of those warehouses are more closely tied to daily fluctuations on the exchange than are warehouse stocks of any other exchange. It may lead to occasional squeezes in deliverable supplies, but ultimately it keeps the exchange in the forefront of the market’s mind at all times.
This hasn’t been lost on Nybot. Remember the eCOPS system? That stands for Electronic Commodity Operations and Processing System, and is a web-based physical commodity delivery system for documentation and data related to coffee and cocoa deliveries. The system has been built to support other commodities as well but is being administered by an open outcry exchange.
Nybot chief financial officer Steve Bass sees no paradox. “eCOPS is not about trading,” he says.
“It is about making it easier for participants on the actual creation and logistics side of the market — as opposed to the buy-and-sell side — to get loans from banks because of the safety of the electronic document. Information flows are faster, more certain and cheaper. But the existence of eCOPS doesn’t change peoples’ minds about strategic positions or strategic philosophies.”
There are lots of other things the system doesn’t do: it doesn’t add transparency to the market on the whole. “The information on eCOPS is owned by the title holder,” he says. “When a bank takes a title to goods in a warehouse as collateral against a loan, they become the title holder, although the product is not really theirs. So the only people with access to the content are the bank plus the beneficiary.
It all still comes down to who knows what and who knows who, just like in any business.”
Other non trading benefits: recoverability. “The information about certified stocks in New Orleans is coming from eCOPS,” he adds. “If we hadn’t implemented this system, you would be dealing with paper records that may or may not have survived.”
Beyond the exchanges
Further down the financial supply chain you find operations like Bolero, which is a neutral secure platform enabling paperless trading between buyers, sellers and their logistics service and bank partners. Their Web site declares, “Our solutions integrate the physical and financial supply chains, providing visibility, predictability, accuracy and security. This delivers improvements in operational efficiencies and reductions in working capital.” Among the services are common legal framework and set of standards enabling the creation of a hub-and-spokes model instead of bilateral, paper-based links. This enables companies to synchronize their supply chains and predict their cash flow.
One example is Columbia’s 560,000 coffee growers, who operate under the umbrella of the National Federation of Coffee Growers of Columbia (FNC). They have been engaged in a rapid process of automation since the late 1990s, beginning with SAP inventory management systems and then moving to Bolero technology. FNC commercial division sub-director Miguel Abisambra says that by reducing paperwork, they have slashed fraud and errors and sped up the entire process. “We had to produce multiple formats of documents and then carry them around town collecting signatures and seals from government agencies and carriers,” he says. “The documents were then couriered to our collection banks abroad, who checked them and couriered them to the customer’s bank. Finally, the customer’s bank forwarded the documents to the customer for payment. The whole process took around three weeks, was costly and diverted our customers and staff from more productive assignments.”
Now, customers can use SAP to confirm their purchases and the Bolero system uses shipping and transaction data to do the rest. As a result of such advances, a global commodities spine is taking shape, making it easier and easier to capture data and distill it into tradable, executable prices. “We are seeing huge advances in being able to quantify risk exposure at every step of the process,” says Triple Point’s Leck. “You really have to be in the business to appreciate how real-time all of this is becoming. Little things — being able to identify transport losses when they happen and file claims — all of this makes the entire process that much more efficient, removes a bit more uncertainty.”
Though relatively unknown among the trading community of the legacy exchanges, companies like Triple Point and Computershare are as visible on the new breed of exchange as are the trading pits on the old. Trading in Freight Forward Agreements (FFA) surged 400% last year, and the Baltic Exchange, which trades freight futures, credits Triple Point with being among the providers making that possible. One reason: such rates are not only dependent on a wide variety of variables fluctuating in real time, but they impact hundreds of cost centers in various users’ books. By offering real-time transmission of information woven into real-time execution, Triple Point enables true hedging of freight costs in a way inconceivable in an open outcry environment.
The Next Generation
Europe’s first generation of trade-matching engines have aged to the point that most customers say they can’t simply be upgraded anymore, they have to be completely rebuilt for the next generation of intra-asset class trading. India’s MCX has launched a new generation of java-based trade-matching software designed to handle smaller markets where the line between hedger and speculator often is blurred. Soon MCX will offer continuous, real-time trading seven days a week via mobile phone. Although such systems are certainly user-accessible, it remains to be seen whether they can handle the kinds of thunderous volumes mature markets generate. Jessup points out that the International Securities Exchange’s huge success in the stock options market came from offering a faster system, so he is sticking with Unix-based solutions written in C and C++.
For retail traders, several platforms offer innovative products that would never exist without electronic trading. The Oanda currency platform, for example, offers a “box option” function, which enables users to draw a box around a price range within which they are executed. Online platform HedgeStreet has gotten attention for its “Hedgelets” — quirky options that offer payouts based on whether they are above or below a certain price. They offer both a binary version, which offers fixed payouts of zero or everything, and a variable version, which offers payouts that depend on how right you are.
However, both products have failed to attract waves of legitimate hedgers. That means no liquidity and as a result, no ability to offer payouts in line with the risk taken. Theoretically, “found” trades will fare better — coming to light as they do in the glow of implied orders generated by thousands of algorithms formed independently but upon the same reality.
“Makes sense,” says Contango’s Furness, “if we allow these markets to evolve as they should and if we truly want them to be recognized as a legitimate asset class.”