Trading places

It is often questioned whether the move by traditional electric companies into the communications business is a worthwhile venture. Junior Isles looks at a recent development by Enron which could revolutionize the comms market and increase the potential for utilities to earn from the business.

Junior Isles

Since the EU directive for electricity market liberalization, there has been news of some electric utilities pulling back from telecomms ventures to refocus on their core business. In the US, however, it`s a different story. For some time, major utilities have been leasing or selling excess fibre capacity to telephone companies and providing communications services that other carriers already offer.

Now, they could soon be moving to the next stage of generating revenue from their existing comms investment through “bandwidth trading”. If it takes off, the idea could revolutionize the comms business, while making it more lucrative for utilities and other potential players.

The concept is being pioneered by Enron Communications, a subsidiary of the global gas and electric company, Enron. Explaining the significance of the concept to the power industry, Tom Gros, vice president of global bandwidth trading at Enron Communications said: “One of the key benefits of participating in the bandwidth trading market is that it will help electric utilities determine whether setting up their own fibre networks is a good investment or not.”

The market would also give utilities the ability to hedge any investment risks. If they planned on making the capital investment to build the network, having a liquid market would allow them to pre-sell the capacity of the network. “If I know it`s going to cost me $20 million to build a link and I could pre-sell it for $40 million; I am probably going to make that investment. I can therefore use the market to remove most of the price risk since I know I will capture a $20 million margin,” Gros pointed out.

This is not unlike the merchant power plant producers who use the market prices of power to determine whether they should build a power plant. However, bandwidth looks more like transmission since it more relates to the right to use a pipeline.

Bandwidth trading could be used by geographically broadly based power companies with multiple facilities that need to be linked. Generally speaking, bandwidth trading will appear as an overlay on the existing telecomms infrastructure. This means that market players will not have to invest significant incremental capital in order to participate in the market.

Expert traders

Enron Communications essentially started out as a subsidiary of Portland General Electric. Like many utilities, especially US utilities, Portland General had a communications group that linked its own facilities within the area of Portland, Oregon.

It was formed as a consultancy to Portland General in 1996 when the act was passed allowing electric companies to get into the telecommunications game. Shortly after, Portland General was bought by Enron. Since its formation, Enron Communications` business has been profitable mainly from the sale of “dark” fibre i.e fibre in the ground without the electronics attached.

Before beginning its venture into telecomms, Enron first did an evaluation of the technologies in the market place and took the decision to build a pure Internet Protocol (IP) network. This was based on the exponential growth of the data communications market. “The average use of the internet in 1990 was to transmit something like a 20 word message. In 2000 or 2001 it will be to transmit live broadcast quality video,” said Gros.

Enron claims to be the first company to build a pure IP network. Explaining the decision, Tracy Smith, marketing communications director for Enron Communications said: “We saw this as the future of telecommunications with IP being the technology that all applications are converging on. It was a less expensive way to build a network, both from a capital and operations standpoint. It also allowed us to focus on the areas that we wanted to focus on i.e. data communications.”

An important advantage which Enron clearly had over other electric companies was that it had no stranded costs or facilities that could hamper its decision when considering what would be best for the future. “This allowed us to enter the market with a clean slate,” commented Gros.

Enron`s ability to start up bandwidth trading stems from its history as an energy trading company. “In the 1980s Enron pushed for a deregulated natural gas market. It was the first company to implement a deregulated mode of operation even before the US was there as a country,” said Gros. Four years ago, Enron had not traded a single megawatt of power; today it is the largest mover of power in North America. Enron was among the first to start trading coal as a commodity and was among the forerunners of the weather trading business.

“The expertise used for trading any of those other commodities [electricity and gas] – whether it relates to the actual trader; or the culture that allows those traders to operate, or the back office systems; and the research systems – are quite synergistic with bandwidth trading,” noted Gros.

“It is yet another commodity that Enron trades. We are leveraging on the skills that we have built up in other commodity markets to expand the portfolio,” he added.

A new approach

Bandwidth `trading` has essentially been happening for years. Typically two parties would sit down and negotiate a contract; sign it say three months later and this would last for a number of years.

As a new entrant Enron wanted to be an efficient player and felt this was not necessarily the most efficient way of doing business. In January this year, the company therefore looked at how it could create a commoditized version of bandwidth trading.

It spent the first three months listening to nearly all the major telecomms players in North America to find out how business was done and what was necessary. The company then decided to unveil a proposal in New York in May this year that would allow efficient trading of bandwidth to take place. “The deal that used to take two to three months or more to close, now takes ten seconds,” said Gros.

Enron sees its approach as very different to what others have tried to do in the past. There are a number of internet brokers who largely broker the minutes market of bandwidth. This is the lowest end of bandwidth. Enron conversely, has concentrated on the higher bandwidth end.

For this concept to work, the first thing Enron has to physically do is agree quality of service (QoS) standards to establish a benchmark for certain “flavours” of bandwidth. Gros explained: “In crude oil for example, we have Brent crude oil in the North Sea which is the benchmark for crude oil produced in this area (UK). Other crude oil extracted can then be sold at discounts or premiums relative to that benchmark.

“To set the benchmark you have to measure the quality of the crude oil. For example, there is a specific gravity and level of sulphur content that has to be adhered to. We are now setting these same types of standards in the US for telecommunications – specifically reliability, latency, jitter and restoration time. ” (see box).

Enron has proposed two city pairs it plans to start with as benchmarks. They are: a time division multiplexing (TDM)-T1 (digital transmission link) service between New York and Los Angeles. With a bandwidth of 1.544 Mbits/s, this will target large multinational bandwidth consumers like Fortune 500 firms. The second benchmark is an IP DS-3 (digital signal) between Washington D.C. and San Francisco/San Jose. With a bandwidth of 44.736 Mbits/s, this will target telecommunications companies and internet service providers.

“These were the two most logical contract sizes – both for large consumers of bandwidth as well as producers of bandwidth,” commented Gros.

One reason for choosing east to west coast city pairs was to provide logical launching points to Europe (from the east coast) and to Asia (from the west coast). “We think it likely, as with crude oil, that bandwidth will evolve quickly to have three or four benchmarks. Once we define the North American benchmark, I see the quick development of an Atlantic benchmark, say from New York or Washington to London. There could also be an Asian benchmark, say from California to most likely Tokyo or possibly Beijing,” said Gros.

Enron sees the European benchmark as critical. Establishing it, however, will have its difficulties. One possibility is establishing a London to Frankfurt route as a benchmark. Gros explained that this is not as obvious a choice as the US benchmark since the real usage of bandwidth in Europe is very distributed. “In America it is very easy to select an east coast-west coast benchmark. In Europe it is more difficult to select the benchmark that would be most useful to all the players,” said Gros.

The next physical task for Enron is to provide connections. “Telecommunications looks like an infrastructure which grew out of independent monopolies not wanting to interconnect with each other. We therefore have to provide connections,” said Gros.

These connections will be provided in key locations where the players can connect their networks on a real-time quality-measured basis. These connections will be called “pooling points”.

Enron is working with Cisco and Sun Microsystems and others to develop the hardware to make the concept work. Cisco will provide the gigabit switch routers (GSRs), while Sun Microsystems will provide the Enterprise servers and the Solaris operating system. In a typical pooling point architecture, there are multiple GSRs interconnected via high speed links forming a pooling point network. The network supports connections for a large number of service providers (SPs). Each SP connects to the pooling point network via one or more links. Bandwidth providers typically connect to the pooling points via very high speed links, while bandwidth consumers connect to the pooling point via lower speed links.

Enron will build the first few pooling points. It hopes that competitors will then “cookie cut” the basic design in other cities. Enron will pay for the first pooling points to be built.

Although Enron is working to establish the pooling points, it will not manage and maintain the system. “To have a level playing field, you need to have a neutral third party to operate the system,” said Smith. Pooling points will therefore be managed by a Pooling Point Operator (PPO), an independent third party responsible for scheduling bandwidth connections, monitoring the QoS of each transaction, issuing code numbers for access to web sites and maintaining the physical security and operational integrity of the transactions. Enron is working with PriceWaterhouseCoopers to set up the PPO.

The PPO will also collect an administrative fee for running the pooling points and to compensate Enron in the first couple of years for its initial investment. This “nominal” fee would be collected from anyone using the pooling point.

How it would work

The bandwidth trading market will be a professional, physical forward delivery market. It is an “over the counter” market where buyers and sellers will call each other directly. Gros explained the set up: “Instead of having to negotiate a one-off contract they can ask: `where are you for benchmark bandwidth for January 2000 to March 2000?` – a three month stretch. If the buyer finds the sellers price to be the lowest offer around, he would agree on the deal.

“The seller would then typically fax a one-page document within one business day to the counterparty agreeing the commercial terms of the trade. It is also likely that same day the buyer and seller would call the PPO and tell them for example, `I just bought one DS3 benchmark contract for January through March, please give me an IP address` which is basically a meter number or manifold point where this will take place. The PPO will issue the IP address and a code number to allow only the buyer and seller to monitor in real-time the performance of that transaction.”

This type of operation will move the market to monthly increment trading as opposed to the market now where deals are talked of in terms of years. Explaining the reason for current multi-year contracts, Gros said: “The reason it is multi-years is because it is very expensive. I would have to invest in documentation and nail up this specific circuit just for this purpose. I am not going to do this just for a multi-transaction. Having this [new] level of liquidity that the pooling point allows, I can now logically ask the question: why not just do it with a monthly trade?”

The scenario is very similar to what is happening in the electricity spot markets. This comes as no surprise to Enron: “The electricity market model is not a bad one. It also comes as no surprise that PriceWaterhouseCoopers, who helped establish the power pooling points in the UK was one of the first counterparties that we spoke to concerning the responsibilities of the PPO. Clearly their experience in the electricity market will be quite useful.”

A DS3 contract is likely to initially trade at about $35 000-$40 000/month. The T1 is expected to trade at about $5000/month. However, Gros admitted: “No one knows what the true price will be because there has never been a market. The price will be determined by supply and demand in the market.”

So how is this infant market likely to catch on? “The first six months of the market will likely be dominated by the producers of the commodity – and that is true of any commodity. There will probably be between ten and 20 major telecomms firms in the market in those first six months,” said Gros. Once the mechanism is proven by the producers other players like some of the large Fortune 500 firms i.e. huge consumers of bandwidth, as well as big financial organizations will probably enter the market.

Enron hopes to have the first couple of pooling points operational by the end of this year. “This would be a great accomplishment considering we only started discussions with the industry in January and rolled out our proposal in May. To be operational by the end of December would be unprecedented,” noted Gros.

With a market up and running so soon, it will not be long before US electric utilities, and eventually Europe and Asia, have another way of making their comms investments work for them.

Comms talk

•Bandwidth: The size of a communications channel. If pictured as a pipeline, it would be the diameter of the pipe. It dictates the speed of data transfer.

•Internet Protocol (IP): This is the most important of the protocols on which the Internet is based. The IP is software that keeps track of all the internetwork addresses and is responsible for routing individual packets of information to their final destination.

•IP DS-3 (Internet Protocol, Digital Signal, Level Three): An IP DS-3 is a digital pathway through an IP network capable of supporting a sustained data transmission rate of 44.736 Mbits/s.

•Pooling point: A pooling point is defined as a switching and interconnection facility in a selected physical location through which connections between bandwidth buyer and seller counterparties may be established, monitored and otherwise served.

•Latency: Technical jargon for waiting time or time delay. It is the time it takes to get information through a network.

•Jitter: Variation in latency.

•Restoration time: The amount of time to restore service after an interruption.