By Khali Henderson

Posted: 09/1999

Is There a Future in Forward and Derivatives
Bandwidth Trading?
By Khali Henderson

The commoditization of
bandwidth as a concept is not new. It is at the core of debate about the future supply and
demand for global communications, and also of the business plans of numerous web-based
matchmaking services for bandwidth buyers and sellers. This summer, the discussion has
moved beyond "what if" to "how to." If successful, such practical
implementation would lead not only to efficient trading of forward bandwidth contracts,
but eventually to a true anonymous commodities exchange, complete with financial
derivatives.

A Modest Proposal

Among those pushing the commoditization of bandwidth is Enron Communications Inc.,
Portland, Ore., a company that is building a national pure Internet protocol (IP) backbone
in the United States that it calls the Enron Intelligent Network, and that claims to have
10,000 route miles of fiber laid. Known more for its first name than its last, Enron
Communications’ parent’s leadership roles in developing U.S. commodity markets for natural
gas, power, coal and weather derivatives is fortifying the company’s proposal to commence
trading bandwidth as a commodity.

In its proposal, which was first made at a May 1999 risk management conference held in
New York and a month later at a similar event in London, Enron outlined in detail
bandwidth commodity offers between two city pairs. The first is a time-division
multiplexing (TDM) T1 service between New York and Los Angeles for large corporate users.
The second is an IP DS-3 between Washington and San Francisco for carriers and Internet
service providers (ISPs).

Enron’s 70-page document, including a proposed master agreement for trading terms and
conditions, has gotten the telecom industry’s attention. "People are taking it
seriously," says Tom Gros, vice president of global bandwidth trading for Enron
Communications. "Whether they agree or disagree remains to be seen."

That’s OK with Gros and his Enron Communications colleagues. "Our intention all
along was to make a reasonable proposal, allow carriers to think about it and consult with
their staffs and then come together to arrive at terms and conditions," he says. To
that end, Enron Communications is planning an industry summit in September or October to
which it will invite a handful of "major players" to discuss commoditizing
bandwidth in North America. Since many on the guest list also participate in global
markets, Gros expects that the discussion eventually would concern Europe and
Asia-Pacific.

Enron’s proposal hinges on the creation of pooling points–switching and
interconnection facilities in selected physical locations through which connections
between bandwidth buyers and sellers may be established and monitored. These pooling
points would be operated by an independent third party responsible for scheduling
bandwidth connections, monitoring the quality of service (QoS) of each transaction and
maintaining the physical security and operational integrity of the transactions. The
network architecture for the pooling points is being developed by Enron’s strategic allies
in this proposal: Cisco Systems Inc., San Jose, Calif., and Sun Microsystems Inc., Palo
Alto, Calif. Cisco will supply gigabit switch routers, and Sun will supply the enterprise
services and its Solaris operating system.

Cisco’s parallel involvement in the funding (along with Microsoft Corp., Redmond,
Wash., and Benchmark Capital, Menlo Park, Calif.) of Equinix, Redwood City, Calif., the
first Internet Business Exchange (IBX) allowing ISPs neutral interconnection, fueled
speculation that Equinix IBXs–planned for Washington, New Jersey and the Silicon Valley
as well as 12 more locations in the United States and overseas–might serve as such
pooling points. Gros has pooh-poohed this rumor, and said that, in fact, Enron has engaged
PriceWater-houseCoopers, New York, to figure out how best to implement that part of the
proposal.

Gros is optimistic that trades will begin before the year ends, but concedes that it is
not really up to him or Enron. "What is beyond our control, but what we are working
toward is consensus in the market."

Such consensus may be hard to come by, however, as most carriers are reluctant to
purposefully commoditize, and thus erode the margins of, their core product.

Power generators expressed the same reticence some five years ago before their market
was successfully commoditized. Gros, who comes to the discussion from the trading side and
not the communications side, says it means he and his teammates must do a better job of
explaining the benefits of such a market. By improving the efficiency of trades (reducing
the time from three months to 10 seconds), the resulting liquidity in the market will give
providers the opportunity to offer more innovative services because the risk has been
mitigated, he says.

Among the obvious benefits to commoditizing bandwidth, says analyst Jilani Zeribi,
Current Analysis Inc., Sterling, Va., is that carriers will be able to expand capacity
easily and rapidly for network buildouts. In addition, companies will be able to reserve
bandwidth for one-off events, such as videoconferences. "The real benefit, however,
will come from the applications and services not yet developed that will take advantage of
the inexpensive and easily provisioned bandwidth," he says.

Besides, as Gros notes, "margins for a true commodity always shrink. Providers can
keep their heads in the sand or embrace it."

Is Bandwidth Destined To Be Commoditized?

The assumption that bandwidth is a true commodity is where Enron finds its most
vociferous opponents who believe that bandwidth does not have the characteristics of a
commodity. One of those companies, Tulsa, Okla.-based Williams Network, also is the
communications subsidiary of a U.S. energy giant, Williams.

"I don’t believe this is appropriate for our market," says Ron Harden, vice
president of marketing, Williams. "We are not going to stick our head in the sand. We
at Williams have the ability to develop the software and systems for a commodity exchange,
but we have chosen not to. When and if the time is right, we will be there."

Harden says one of the primary reasons he dislikes the idea is that it drives the basis
for competition to price, noting that networks are not created equally. "In
electricity, an electron is an electron. But a DS-0 mile is not a DS-0 mile," Harden
says. "There are many ways to differentiate capacity, such as in provisioning,
redundancy, security, customer service and price."

"Of the six or seven criteria required for a commodity market, almost none are
evident in the telecommunications market," Harden says. Williams commissioned a study
of this question by Boston-based Boston Consulting Group Inc., which determined that a
commodity market requires a fragmented supplier base, fragmented customer base, price
volatility, common unit of exchange and agreed-upon terms, a delivery mechanism and
limited vertical integration.

"A commodity is a commodity," says one of the study’s authors, Bill Cornog, a
vice president in BCG’s high-tech practice based in Dallas. "The question is: Can you
turn something into a commodity? I don’t think so."

BCG’s study found that, while there may be a fairly fragmented customer base, the
supplier base for bandwidth is not, Cornog says. On any given route, there are only 15 to
20 suppliers with a handful having majority control. Most commodities have a more
fragmented base, he says, noting that industry consolidation only will lessen the number
of suppliers.

The study also casts doubt on all the existence of other needed criteria, including the
availability of a delivery mechanism (the required level of intercarrier connectivity does
not exist), as well as on the existence of a common unit of exchange (bandwidth comes in
many different protocols, each having a different degree of robustness).

Additionally, Cornog says that pricing is going down, but that trend does not equate to
volatility. Others agree. "Bandwidth is not a finite resource like oil and coffee,
it’s an infinitely upgradeable service like microprocessors or RAM," says Marcus de
Ferranti, director for Band-X, London, the first web-based bandwidth trading floor.
"As a result, there is no volatility in pricing, so while a bit of forward trading
here and there is both happening and likely to continue, there is no point or likelihood
of derivatives being worthwhile."

On the point of price volatility, Enron’s Gros says it doesn’t make sense to argue that
there is no volatility since the market is not yet transparent. "That’s absurd,"
he says. "In the power market, they used to say pricing wasn’t volatile. When the
commodity market emerged, you can see it. … Volatility cannot be observed unless you
allow it to be observed," he says, noting the chicken-and-egg argument.

And while Gros concedes that a true exchange with financial instruments is a natural
outgrowth of a commoditized market, he says that the Enron proposal does not go that far.
"A commodity market is not a commodity exchange," he says. "An exchange is
not what we proposed."

Instead, he says that trades will not be anonymous and pricing will not be indexed. The
initial routes proposed, he adds, will serve as pricing benchmarks, not unlike in crude
oil where some 300 grades of oil are traded using one benchmark. In this scenario, he say
pricing transparency will occur only among market participants, who–through frequent
trades–come to understand what the market will bear, and not to speculators. "It’s
an over-the-counter market," Gros says. "Buyers will buy at the lowest price so
even if you [as a bidder] were not in the closed deal, you have gained an understanding of
pricing."

The Next Step

Enron is basing its proposal on a market structure of the past, says Ross Mayfield,
vice president of marketing for RateXchange. "It is based on market discovery through
a telephone tree of brokers," he says, noting that this loses the potential that a
web-based trading environment offers.

RateXchange intends to effect bandwidth trading–goods and contracts–using its
web-based exchange and physical network of 12 switching hubs called the Real-Time
Bandwidth Exchange (RTBX). Currently, RTBX only trades minutes in New York and Los
Angeles, but is gearing up to trade other forms of capacity.

Mayfield says RateXchange will launch an industry educational campaign in September to
help carriers understand the exchange proposition. He says one of the biggest stumbling
blocks to success has been carriers’ reluctance to understand its benefits and even how to
do it. "We’re anticipating their readiness and trying to push them along," he
says.

Having received necessary growth capital from its acquisition by NetAmerica.com Corp.,
San Francisco, in June, RateXchange will use this educational period to continue
development on its trading system, Mayfield says.

In conjunction, RateXchange has begun publishing forward pricing indices. Its Revealed
Forward Pricing (RFP) Index, published in August, is a compilation of major East-West
routes (see chart, below). Another player, Arbinet, New York, which is known for its
real-time minutes exchange, also is anticipating the emergence of an anonymous trading
environment by launching in May online trading of bandwidth derivatives–forward contracts
and options–wherein the buyers and sellers are suppliers and consumers are speculators.

Graph: DS-3 Bandwidth RTBX*RPI and RTBX*RFP-360 Indices

In a statement, Arbinet states forward and option derivative trading allows companies
to hedge their exposure to future moves in the price of capacity. Furthermore, the company
says it provides a level playing field where even small operators can realize the pricing
advantages, which were, until recently, available only to large incumbent
telecommunications carriers.

"Arbinet aims to be the premiere provider of risk-management solutions for the
bandwidth market," says a spokesperson for the company. "As a nonproprietary
player, we can offer an impartial executive for our clients, which is important in a
nontransparent and illiquid market like this."

Staged in two phases, the new exchange at www.ebandwith.net initially will offer a
trading floor for anonymous bids and offers popularized by Band-X in Europe and
RateXchange, San Francisco, in the United States. In phase two, beginning at the end of
the year, Arbinet expects to move to a physically settled environment wherein carriers
connect to its hub switch, and traffic is routed automatically based on set criteria and
credit policies. The company is in testing with several customers with routes to Brazil
and Hong Kong.

What is promised in phase two, says Arbinet President and CEO Alex Mashinsky, is a
reservation market wherein carriers can make short-term, forward-looking contracts. For
example, a carrier that has purchased an indefeasible right of use (IRU) on an undersea
cable could lease out some of that bandwidth for a shorter time period than its own
25-year contract. That way, the pipes are filled until the time when the carrier needs it
for its own traffic.

While several exchanges could develop simultaneously, in the end, only the largest
would have the greatest liquidity. Thus, only one or two global exchanges could survive.
That is, if they ever get off the ground in the first place.

Energy All Over Again

The commoditization of bandwidth is at its earliest stages, if at all. Like Enron, many
proponents like to draw parallels to energy markets as justification for their predictions
and pressure toward this outcome. To those carriers that say, "Our market is special;
our product can’t be stored," Mike Moore, an energy broker with Amerex, Houston,
says, "It’s happened in power."

The power market went from a monopoly industry to a regulated commodities market in
little more than five years. As a result of the 1992 Energy Policy Act (EPACT)
deregulating the electric power industry, there are an increasing number of commercial
products and merchandisers, including more than 450 power marketers that buy and sell
energy and capacity, but do not own generating or transmission facilities. The transition
to competitive, determined prices for wholesale power has created a corresponding need for
effective price risk-management instruments, writes Paul J. Patano Jr., a partner with
McDermott, Will & Emery, Washing-ton, in an article in the April 1999 issue of Futures
& Derivatives Law Report. During the past two years, Patano reports the spot or cash
market for wholesale power has increased by more than 200 percent, and market participants
increasingly are using options to obtain and sell power. Furthermore, eight regional
transmission hubs have emerged as trading interconnection points. As of February, he notes
that the CFTC has approved 12 electricity futures and options contracts.

Among the parallels drawn between power and bandwidth are three indicators–regulation,
market demand and technology–of the inevitability of a commodity market, says Enron’s
Gros.

He reasons: In power, the Federal Energy Regulatory Commission (FERC) Order 888 in 1996
led to the deregulation of the power industry; in telecom, the Telecommunications Act of
1996 is opening markets to competition. In power, market demand was not well-matched to
supply, in part because of regulatory restrictions creating pent-up demand; in telecom,
demand for bandwidth for multimedia applications continues to grow. In power, the
introduction of efficient gas turbines made possible the economic production of power for
peak production times; in telecom, IP has emerged as a convergence layer joining voice and
data for more efficient transport.

Also mindful of these parallels, Amerex’s Moore is among many in the energy industry
that are watching developments in the commoditization of bandwidth with an eye toward how
they can be involved. For his part, Moore intends to make a role for Amerex in much the
same way it has served the power industry: as a go-between for buyers and sellers. For him
the fit is natural: not only are his energy clients bandwidth buyers, they are sellers.

And, he warns, they are experienced traders, having five years up on bandwidth
carriers. "One of the biggest arbitrage opportunities is market ignorance,"
Moore notes of the energy companies’ trading experience.

Moore understands the reticence among carriers to embrace commoditization of bandwidth,
recalling the same sentiments expressed by power generators about electricity. In the
beginning, many will be unwilling, but it only takes one or two to break ranks and force
the issue, he says.

If the electricity market is indeed an appropriate model, the evolution to a commodity
market begins by defining the product–the stage at which bandwidth is now, Moore says.
From there, he says, will come the creation of hubs for service delivery, which, in turn,
prompts the emergence of independent system operators (ISOs) to manage those hubs. Then
there will emerge multiple futures contracts and multiple exchanges run by neutral
parties. Finally, at the urging of the financial community, audited pricing indices and
financial derivatives will appear.

The emergence of a purely financial market for bandwidth will emerge when certain
conditions are met, according to a presentation prepared by Enron. There needs to be
sufficient history to perform meaningful statistical analysis to determine risk parameters
and allow use of familiar risk management tools. There also needs to be an unbiased price
index reflective of the physical market, as well as liquidity sufficient to attract
speculators so they, in turn, increase liquidity, and creditworthy counterparties to
mitigate risk.

For now, the financial community is watching from the sidelines with a curious eye. No
formal investigative committees have been formed, but the opportunity, they say, is
definitely on radar.

Khali Henderson is editor-in-chief of PHONE+ Magazine.