Trading places

15 November 2010 |


Eira Hayward looks back at the attempts to engage the industry in capacity trading a decade ago and forward to the commodities of the future cloud era.

Only a few old lags will remember, but Capacity magazine in its early life was “the new magazine for the new market in trading telecommunications capacity”. It was a guide to take the industry into the brave new world of trading capacity as a commodity product, like coffee, pork bellies or orange juice, complete with exchanges, spot markets, pricing indexes, futures options and the rest. It would work, we all said. It had been done successfully with other utilities like electricity and gas; all that was needed was a bit of education and then everyone would join the party.

By 2000, the energy companies were wading in and big carriers were setting up trading desks. Enron, then one of the world’s foremost energy companies, made huge amounts of noise about its role as a market maker, citing its vision in recognising the opportunities to transform the market when natural gas was deregulated in the 1980s as all the evidence anyone would need to show that it was right to take the same approach with telecoms.

In February 2001, it was estimated there were 25 telecoms exchanges, and analysts believed that carriers would have to bow to inevitable pressure and start trading bandwidth. Ovum produced a report setting out why it thought bandwidth trading was inevitable and would transform the telecoms market; an industry body called the Bandwidth Trading Organisation was formed; and Comptel had a working party formulating a “master agreement” that would serve as a standard form agreement for the trading of bandwidth units.

Liquid digital markets

The early issues of Capacity make for fascinating reading and underline how dramatically different the telecoms industry was in the early years of this century. In early 2001, Yankee Group conducted some research with Capacity to look at what was needed to make bandwidth trading widespread (see graph on page 54, A scorecard for liquid digital markets). Capacity looked at the drivers and barriers to making this liquid digital market a reality. In its favour was market demand; counting against it were entrenched attitudes, bilateral agreements, no standardised contracts, lack of regulation in the trading market, and the need for local loop unbundling.

Says Stephan Beckert, research director at Telegeography: “A lot of people drew parallels between the liberalisation of the telecoms industry and the deregulation of electricity that had already happened. They thought they’d have a business in trading, just like they did with natural gas and electricity. But circuits are not identical, and there’s no easy way to trade them. A lot of it was hype that never made sense.” Beckert says that Telegeography was even approached at this time by the Wall Street Journal and asked to develop a bandwidth spot market index. Although his company declined the request as it couldn’t see a future for bandwidth trading, he says that it “was hard to stay out of that hype”.

But in December 2001, Enron filed for bankruptcy. Its name has since become a byword for corporate corruption and its subsequent unravelling brought about the dissolution of accounting firm Arthur Andersen. It was also the beginning of the end for the idea that capacity would be traded widely as a commodity product. Says Beckert: “We never found any carrier who had traded a circuit with them [Enron] or bought a circuit from them. They were arrogant, and, as it turned out, they were lying.”

Matthew Finnie, CTO at Interoute, concurs. His company was one of many that Enron failed to sway with its vision of bandwidth trading. He remembers: “We always felt we weren’t part of the cool gang. But a lot of the people talking about trading bandwidth back then were clueless. What struck us was the need for liquidity in the market. There wasn’t enough activity. It was hardly Wall Street.”

Market conditions

Camille Mendler, vice president of Yankee Group, says that although criminality was rife in the Enron era, the market conditions and technologies at the time were wrong for commoditising digital assets. For a commodity market to be successful you need liquidity – enough people willing to trade – and what the commodity markets like to call fungibility – or standardisation, a standard unit for sale. As Mendler points out: “Carriers don’t like to think their products are interchangeable.”

But she understands why so many thought there was a fantastic opportunity. Over 1,000 new telcos were created in the three years following EU telecoms deregulation in 1997 so surely they would want a centralised means of trade. Networks were being built out like mad, and there was plenty of money around. “But 10 years ago there weren’t enough buyers, exchange points were not globally available and there were no standard contracts. Enron was trying to create a market for which there just weren’t enough people who were similarly bullish,” she says. There were other reasons too – a glut of capacity and a downward spiral in prices. “Why would you have hedged against pricing and future shortages?” asks Beckert. “There was complete predictability about pricing and availability.”

Apollo MD Richard Elliott, founder director of possibly the best-known bandwidth trading exchange, Band-X, comments: “Why didn’t bandwidth trading work as a market? That would take a 10,000 word article to explore in its own right. There’s sometimes talk about trying to resurrect it – and I’m happy to give anyone the benefit of my experience. At the time it seemed like a logical extension of trading in other utilities – like oil and energy. I remember Enron did due diligence on us – and then went off and did it themselves. That gives you an insight into what sort of company they were. Eventually they were making all sorts of claims about the future of telecoms trading that it was clear to me were hard to believe in. But it was difficult for us to point out that the emperor had no clothes on, as we were still in the process of evaluating a new direction for the business. The market didn’t feel much like a party at that point. I was lucky in the period after 2001 to recruit some excellent colleagues, some of whom I worked with up to the sale of the company. That influx of skills was a major factor behind the success we had.” 

Try it again 

Though we still have the bilateral tables and opacity of pricing of old, as Elliott comments, the concept of tradable commodity telecoms products has never entirely gone away. Periodically someone has another crack at trying to make it work in some shape or form. Just in the last couple of years we’ve seen the formation of companies like Nytex, or the New York Telecoms Exchange, and the online escrow company Vertecto, set up to make it easier for carriers to do business with each other through exchanges. Equally we’ve seen Buysellbandwidth, the brainchild of Flag Telecom’s Neil Tagare, seemingly founder this year. The attempts of Global Capacity to produce a pricing index and inject some transparency into pricing also faltered when it entered Chapter 11 bankruptcy protection in the summer.

Think a little harder and there are a sizeable number of companies still trading and exchanging telecoms in some form or other. Certainly voice minutes trading has never gone away. The biggest carriers have voice trading desks and their own closed exchanges; they may be buying a lot of voice termination for their own business needs, but they also sell a lot too.

At Nytex, CEO Lawrie Taylor-Deutsch is a happy man. Launched 18 months ago as “the world’s first neutral international telecommunications commodity exchange”, the company has had its sceptics from the start. It has found a role in trading voice minutes in emerging markets, “high-value markets”. “We work with termination rates north of 5 to 6c and as high as 30c,” says Taylor-Deutsch. “Our model doesn’t work at under 1c markets. Volatility is not a key to success for us. We don’t see people buying to hedge three months out; they’re buying for their immediate use.” Taylor- Deutsch says that as well as routes in Africa, Nytex is finding good business in European mobile termination. Although he’s coy about disclosing any numbers, he says the exchange is meeting expectations and its year-end figures were a pleasant surprise. “Our strength is voice. We’re not trading bandwidth, and we have no immediate plans to look at it,” he says. 

Bandwidth exchanges 

Some of the old names are still with us. Arbinet and Routetrader offer spot markets for telephony minutes. Arbinet, which unsuccessfully formed an IP exchange in 2004, has had its troubles recently. It lost its CTO and senior VP of sales and marketing in the summer in a bid to cut costs, while continuing losses have led it to rethink its strategy and introduce new services.

Says Sue Su, vice president, data services at Arbinet: “Arbinet’s IP Exchange business began in 2004 under the assumption that IP transit would become a commodity and customers would be able to buy internet bandwidth in a self-served automated trading environment (much like the premise of the traditional bandwidth exchanges at the time). We soon discovered that not all ISPs are created equal; hence trading IP bandwidth was not conducive to a pure commodity price play. In addition, we realised that customers had a decreased desire to play the price bidding game once they found suppliers they liked.”

Arbinet has modified the dynamics of its IP exchange trading model to give customers a choice of suppliers with much simpler and more manageable commercial arrangements. Adds Su: “At the end of the day, trading IP bandwidth didn’t work like a traditional commodity or the way people imagined because of the complexities associated with IP configurations and network interconnections.”

Interoute has long had a virtual voice network to support its customers’ voice requirements – and Finnie points out that with Ethernet as a common transport mechanism in networks it is becoming easier to take away some of the objections to bandwidth trading. “Surety of service is important,” he says. “We will see the technology available to allow a breadth of service off the same interface.” An evolution of this then would be the ability to connect with other suppliers, which would bring trading opportunities.  

Exchange points 

Epsilon last year introduced Epsilon Connect, billed as “the largest global meetme room for voice and data connectivity, linking the major regional telecom hubs around the world”. Epsilon CEO Andreas Hipp says: “Our Epsilon Connect platform is actually nothing else but a pooling point where multiple networks aggregate and cross-connect.” People who have large optical ports with Epsilon can cross-connect to another carrier on the platform within as little as 48 hours. “I’m sure this supports buying and selling of capacity but wouldn’t go so far as to say trading. But we are slowly getting to a point where the physical interconnect points are so large – we have already 270 carriers – that buying, selling and physical delivery should be possible in a matter of days.”

This becomes more relevant on expensive and single line cables such as subsea cables. Then carriers need capacity fast and only for a month or two, and “the transaction becomes pointless if timelines are not met”. Epsilon performs such transactions as it has the exchange points and the carriers. “If they have no direct commercial agreement between each other, we step in,” says Hipp. “We try to promote the idea and the scheme, but still it’s something which will take time to be accepted and understood by many parties in order to see volume. However we have a few good scenarios and will build on that, but without any fancy databases and a trading platform attached. This might follow if we see better acceptance.”  

Future intent 

Widespread commodity markets for telecoms will be upon us before we know it, say some observers, albeit in a more subtle way than 10 years ago. The technology has moved on sufficiently that many of the old difficulties of making bandwidth a commodity market could disappear.

Mendler points out that cloud marketing literature often borrows jargon from the commodity markets – liquid, spot pricing, exchange and so on. “Is this just marketing speak or a signal of future intent?” she asks. Google received US regulatory approval to enter the wholesale energy commodity trading market in February. “Do linkages exist between energy commodities, cloud computing and other digital assets?” asks Mendler again. She points to Amazon, which runs a spot market for its own virtual machine instances, and Deutsche Telekom’s Zimory, a marketplace which brokers cloud infrastructure services, as the forerunners of a possible new world order. Then, of course, there are all the Ethernet peering exchanges. Mendler also sees a possible future in brokering content. But she adds: “It’s important for industry players not to repeat the costly mistakes of the past. Making markets warrants caution, even if you are the smartest guys in the room.”