Deception, or Just Disarray, at Enron?
The New York Times
June 8, 2003
By KURT EICHENWALD and JOHN MARKOFF
AFTER more than a year of investigation, the collection of millions
of pages of documents and interviews with an untold number of witnesses,
federal prosecutors last month brought their most sweeping fraud indictment in the Enron case,
charging a group of executives with illegally
pumping up the company's stock price by lying
about the existence of certain Internet technology.
But now, whether anyone charged with those crimes will ever go
to jail may rest in part in the hands of an unlikely group of people, including Drew
Carey, Shania Twain, Pete Sampras and some fans of independent short films.
The reason is simple. The defense is expected to argue that
Enron possessed - at various levels of quality - the technical ability it claimed. And
the proof, say software specialists from Enron and other industry
executives, lies in part in the technical features that the company offered while
streaming video over the Internet for Mr. Carey's comedy show,
the 1999 Country Music Awards, the Wimbledon tennis competition and a Web site
that provided an assortment of independent short films.
Indeed, for all the seemingly black-and-white fraud allegations that
have roiled Enron, the charges involving technology deception may present far
more gray. At bottom, the technology in Enron's Internet division - Enron
Broadband Services, or E.B.S. - may have been more developed than the
indictment suggests, presenting prosecutors with some complex challenges.
That is the conclusion drawn from interviews with an array of the unit's
former customers, suppliers, consultants, employees, partners and
competitors. Many of them argue that the charges about technology
reflect not criminal activity but the disarray and disappointment that typically
accompany product development in the computer industry.
In that world, claims are made for a technology's abilities long before commercially viable
products are available - and, sometimes, results fall short of hopes.
"If they succeed in convicting the Enron developers," said an
executive at a major computer hardware manufacturer who was never employed by
Enron but who had direct knowledge of its systems, "anyone
in Silicon Valley can be sent to jail."
Former Enron executives voice similar opinions. "I do not think
the technical capabilities were overstated," said Larry Ciscon, former vice president
for software engineering. "And comparing E.B.S. to the other software
companies I've seen in my 15 years in the software industry, I did not see
anything outside of the standard product-development process."
Prosecutors, however, say they have no doubts about their charges.
"We have full confidence that the proof at trial will bear out the allegations in
the indictment," said Leslie Caldwell, head of the Justice Department's Enron task force.
None of this means that the Enron broadband unit was well run
or that its business prospects were not exaggerated. Executives may have boasted of
technological capabilities long before they were available, and
financial illegalities could also have taken place. Indeed, large portions of the
broadband indictment against seven former E.B.S. executives pertain to
accusations that the division manipulated its reported income through the
use of the same types of off-the-books partnerships that played a huge
role in Enron's collapse. Such dealings led to criminal charges against several
executives, including the former chief financial officer, Andrew S. Fastow.
Moreover, there is plenty of the type of evidence that might convince a jury
that insiders knew something was amiss inside the broadband division.
The executives charged in the criminal case sold millions of shares
of Enron stock as the price was soaring over excitement about broadband. That
has led to criminal charges of insider trading, with the government
arguing that the executives lied about the division's capabilities for the purpose
of pumping up the stock price so they could dump their shares.
For that accusation to stick, however, the government will have to prove not only that
the statements were untrue, but that the defendants knew it.
But the primary charges against four of the seven defendants - Kenneth Rice,
Joe Hirko, Scott Yeager and Rex Shelby - state that certain
software-driven functions promoted by Enron as available on its
network were not. Prosecutors contend that Enron executives falsely said that the
company had products for shipping and streaming video and other media
over the Internet to corporations, and that it could meter and bill customers
based on use, as well as allow customers to select qualities of service and
schedule times for shipping data across the network.
In addition, the government argues that Enron's related software initiative,
known as the Broadband Operating System, or BOS, did not work as
Enron claimed. The initiative was intended in part to give outside
software developers access to the features of the Enron network.
Interviews with customers and others suggest that the functions were
available, although some in more rudimentary form than others. Indeed,
prosecutors' ability to prove the case, people involved in the
investigation say, will turn on subtle questions like whether a company can claim to offer
a function based on crude versions of software that may not be marketable on a large scale.
Technology efforts like Enron's broadband venture have routinely begun
with a grand vision to which initial commercial products too often fail to
measure up, particularly in their early releases.
Indeed, version numbers of commercial software have long
been a running joke among customers who have come to expect that bugs and
shortcomings will not be ironed out until the second or even third major release.
For example, it was not until Microsoft, the world's dominant
maker of software, introduced Windows 3.1 that many industry
analysts and executives felt that it had a competitive product, despite grand claims
made nine years earlier when the product was announced.
There is no doubt the government has obtained evidence showing
that Enron's broadband division was deeply troubled. In an e-mail message dated
Dec. 20, 2000, Bill Collins, a former director of business development, lamented
that Enron's network software effort was fizzling, with no market
share, no purchasers and no one using it. "I don't care what lipstick
and rouge you paint that bitch up with," Mr. Collins wrote days before resigning
from the company. "She's still just dead meat lying on the sofa,
just threatening to stand up and steal the show."
But at the same time, programmers and engineers were sending e-mail
messages to their superiors - including several men now under indictment
- boasting about tests of the software functions now at the center of the charges.
Those included tests for clients ranging from Warner Brothers for
its Drew Carey Webcast to countrycool.com, which streamed
the 1999 Country Music Awards over Enron's network.
"The results were successful," Kirk Wright, an Enron engineer, reported
in an e-mail to his bosses on Nov. 10, 1999, about a test of a project to
stream an episode of "The Drew Carey Show" over the network.
He added, "All of the streams worked and started within the acceptable time frame,"
signaling that Enron's broadband scheduling functions had performed properly.
The government indictment suggests that such functions did not
exist.
Moreover, data collected in the 1999 Webcast of the Country Music
Awards show that Enron was metering use of broadband by customers, at varying
levels of quality. These, also, are functions that the indictment suggests did not exist.
Ultimately, many technical experts at Enron made public statements
about the technology that are virtually identical to comments cited in the
indictments, raising the possibility that there were true believers inside the division.
For example, in a Jan. 12, 2000, conference call with Lehman
Brothers analysts, David Berberian, managing director of the E.B.S. technology group,
described the system's abilities in ways prosecutors now say are false.
The network software "fundamentally controls how data is routed from
content providers to users, how quality of service is determined appropriately
by path, and how we channel the data down the right path," Mr. Berberian
said, according to a transcript of the call. Mr. Berberian did not return
phone messages.
Legal experts say that it is unusual for a situation in which people
argue over a product's functionality to result in criminal indictment. That is
largely because of the difficulty in proving that a decision based on conflicting
opinions was knowingly false.
"If a decision was made that the broadband system worked, and
that is based on engineers and computer experts saying it worked, that should be
enough, period," said John J. Fahy, a former federal prosecutor in New Jersey.
"Indictments and criminal charges shouldn't be based on facts that
are the result of different people having different interpretations of
whether something works. That's just not fair."
The events that led to the recent charges involving broadband began
with Enron's purchase in 1997 of Portland General Electric, an Oregon utility.
The new subsidiary came with a start-up telecommunications business,
which Enron initially planned to shut down. But a successful fiber optic
network deal known as the Western Build - in which Enron sold
a small portion of the lines involved at prices that paid for the entire project -
persuaded Jeffrey K. Skilling, then Enron's chief executive, to pursue
the building of a broadband business.
Immediately, the effort descended into chaos, as the Portland unit
and the Houston headquarters pursued different visions of the future. Portland
embarked on an effort to build the ultimate broadband network for the Internet,
involving large-scale laying of fiber optic cable and development of
software. Broadband executives based in Houston, meanwhile,
focused on trading the use of bandwidth to allow companies to ensure they had the
ability to transmit data when they needed guaranteed levels of quality
One strategy involved building a network, while the other focused on
creating a virtual network through trading capacity. In addition, Enron embarked
on efforts to provide content delivery, theoretically increasing
demand both for trading and for its own network.
In essence, Enron was trying to pursue multiple, contradictory
business lines - any one of which would have taken dedicated efforts by the
company - and all within a single division that had yet to prove there
was enough customer demand to sustain any of it. Leading it all were Mr.
Rice and Mr. Hirko, two executives with backgrounds
in the energy business whose limited technical expertise was often the subject of ridicule
within the broadband division.
"We were trying to do too many things at once, and not focusing,"
said one former Enron software programmer who spoke on the condition that he
not be identified. "The whole giant ambition was just too big. But that doesn't
mean that there wasn't a network and there wasn't software. But if
you have a thin market for these things to begin with, and you're
trying to do five things at once, it doesn't all add up."
THE disorganization led to a poisonous culture in the broadband unit,
former executives said, with the only available means for advancement often
being to attack the technologies and strategies of other executives. In that
environment, the technical experts working on building a network
decided in late 1998 to buy a tiny Houston-area software maker called Modulus Technologies.
Modulus, founded in the 1990's, had created a programmer's
tool kit called InterAgent that was used to simplify the building of complex programs
that linked different kinds of computers and operating systems
Before Enron approached, Modulus was negotiating to be acquired by Sun
Microsystems, and Sun executives were stunned when they learned
that Enron had paid $30 million for the company, which had eight employees.
Although Enron deployed InterAgent widely on the computers
that managed its network, the company was slow to take advantage of its potential.
Some programs, like an application called Bandwidth Manager, were built
entirely without relying on InterAgent, according to former software
writers with Enron.
Still, the technical experts at Enron were singing the praises
of InterAgent, proclaiming its ability to provide services that the criminal indictment
appears to contend were unavailable. "InterAgent gives us an intelligent
delivery platform for the services traveling over our network," Stan Hanks,
Enron's former vice president for engineering, told an industry publication
in early 1999. "It also helps us ensure that people are only getting content
they've agreed to be billed for."
Mr. Hanks, who defense lawyers and former Enron executives
said they believe is now a witness for the government, declined to comment.
By that time, the Enron network had apparently become
real enough so that the company could make an aggressive pitch to a small Seattle-based
start-up, AtomFilms, which pioneered the idea of streaming digital videos
of independently produced films to personal computer users.
That year, AtomFilms was looking for a way to show videos to its customers
without the jerky interruptions that had plagued this technology, and
Enron assured the company it could do so by bypassing the dozens
of hops that data made as it typically traveled through the open Internet.
AtomFilms agreed to let Enron offer its movies, and soon the film company
was sending four million to five million video streams a month to its
customers from Enron's video servers at different sites nationwide.
An AtomFilms' executive, speaking on condition of anonymity, said
his firm needed to know details about who was watching which films, and for how
long. So it routinely used an online reporting system that Enron provided,
indicating Enron could offer metering and billing - again, services that
the indictment appears to argue did not exist.
At the same time, Enron embarked on its grand plan to create
BOS - the Broadband Operating System. But rapidly, former employees said, the
entire concept faced an important hurdle of nomenclature: in the chaos that
was Enron Broadband Services, the term BOS was interpreted
differently by various factions, according to former Enron employees and internal documents.
To some, it described functions that Enron performed on its network;
to others, it was a software interface that allowed outside programmers to gain
access to those functions for writing applications; and to still others
it was simply the name of an industry protocol for the Internet. Indeed,
according to one former programmer, the name BOS was created before anyone was sure what functions it represented.
On Jan. 20, 2000 - at an analysts' meeting that prosecutors
contend was part of the effort to deceive - Mr. Skilling announced that the company
had shipped a beta test of BOS that would allow outsiders to develop
applications based on Enron's network software. According to a transcript of the
meeting, he said the complete set of software would not be finished for six to nine months.
IN the indictment, prosecutors contend that Enron never disclosed at
the Jan. 20 meeting that BOS was in development. Asked about the apparent
discrepancy with Mr. Skilling's statement, a government official involved in
the case said that no beta test - in which an early form of a program is
sent to outside software writers - was ever conducted. In subsequent
interviews, however, software writers at major computer corporations said
they indeed conducted such tests for Enron on the functions described by Mr. Skilling.
Over the next year, a shakeout began in the broadband industry. The business
of Enron's primary customers - Internet service providers for
consumers - fell apart, and the company shifted its attention
to business-to-business applications and video on demand. But by mid-2001, the
overall broadband marketplace was in a free fall, and Enron effectively shut down E.B.S.
Now, former programmers and technicians lament the government's
portrayal of the broadband effort as little more than a lie given the patina of
success just to lift the stock price.
"There was more that existed than you would believe from reading
the text of the indictment," one Portland programmer said. "If we had been given
eight months to a year, it would have been amazing."
Copyright 2003 The New York Times Company
nytimes.com |
