In the late 1990s, the telecom map was being redrawn by faith as much as by engineering. Undersea cables, long-haul fiber, and privately financed networks were supposed to dissolve geography itself, and the people pouring money into the sector were rewarded for believing that bandwidth would behave like a new kind of oil. Global Crossing was born in that atmosphere, when the promise of cheap, abundant connectivity seemed to outweigh the practical question of whether anyone actually needed — or could profitably absorb — that much capacity.
Gary Winnick, a financier with a reputation for speed and appetite, stood at the center of that formation. According to contemporaneous reporting and later securities litigation, he came out of the high-yield bond world with a talent for raising capital and persuading people to suspend caution. He was not a cable engineer; he was a deal maker who understood how markets reward scale, narrative, and momentum. That mattered, because the company he helped build needed all three. It was not enough to lay fiber. It had to look inevitable.
The period matters because the market itself was already primed to accept that inevitability. In the late 1990s, telecom operators were expanding on the assumption that bandwidth demand would keep climbing and that capacity would become the basic measure of value. Analysts, lenders, and investors were willing to treat future utilization as if it were present fact. A network that was lit, mapped, and announced could be valued as though customers were already waiting at the other end. In that atmosphere, the distinction between infrastructure and story blurred. The story often moved first.
Global Crossing’s early world was the deregulated, exuberant telecom economy of the era: carriers expanding before demand fully materialized, financial analysts valuing promises more than cash flow, and a market that often treated capacity as if it were already sold simply because it had been lit. The structural weakness was not a single loophole but a systemwide habit of optimism. Intercarrier agreements, bandwidth swaps, and provisioning arrangements were complex enough that outsiders could not easily see whether a transaction reflected real demand or merely rearranged future obligations.
That complexity gave the company room to operate in a gray zone that was difficult for investors to police. The transactions at issue were not exotic in form; they were ordinary enough to fit into the business vocabulary of the industry. Capacity leases, interconnection arrangements, and swaps among telecom companies could all be described as commercial activity. But when the industry’s demand had not caught up with its buildout, those same arrangements could be used to make revenue appear more robust than the underlying customer base justified. The accounting effect mattered as much as the physical network. If a circuit or block of capacity could be treated as sold, then the income statement could show growth before the market had delivered the cash.
A scene from that moment is visible in the company’s physical scale. The firm pushed a global fiber network into service and presented itself as a new backbone for the internet age, with landing points, rights of way, and branded offices meant to signal permanence. In Manhattan and on investor roadshow circuits, the company’s message was the same: this was infrastructure for the next century, not a speculative experiment. The sensory language of the business was hard metal, glass, and illuminated maps; the financial language was margin, growth, and booked capacity. The gulf between them would matter later.
The geography itself added to the illusion. A company with cables crossing oceans and facilities in major markets could appear too large, too technical, and too global to be vulnerable to simple accounting games. That perception shielded the enterprise from skepticism. Scale became part of the seduction: the bigger the network looked, the more plausible it seemed that demand must be waiting somewhere inside it. Yet the same scale also made it harder to see where real customers ended and circular transactions began.
The germ of the scheme, according to later accounting investigations, was not necessarily born in one dramatic stroke. It grew where the business model became inconvenient. Telecom companies had excess capacity, but the market had not matured enough to pay for all of it in cash. Rather than let the numbers show the drag, companies began to exchange capacity, circuits, and future access rights in transactions that could be described as revenue even when the economics were thin or circular. The first line crossed was conceptual: if two firms swapped assets and each recorded a sale, the books could show expansion without corresponding outside demand.
That was the founding lie — not that fiber existed, but that the market was deeper and more profitable than it was. The lie was structurally attractive because it was easy to dress up in the language of industry practice. Executives could say they were simply managing assets efficiently. Accountants could say the transactions had form. Analysts, eager for signs that telecom expansion had an endpoint of profitability, often accepted the story with little resistance. The system did not require a conspiratorial mastermind so much as a shared willingness to see volume where there was only rotation.
The records and later proceedings around the company would focus attention on how those arrangements appeared in the books and disclosures. Revenue recognition entries, capacity swaps, and intercarrier deals sat in the ordinary machinery of corporate reporting, where they could be buried inside broad categories rather than stand out as obvious warning signs. That was part of the danger. The suspicious activity did not necessarily announce itself in a single line item. It could be dispersed across many transactions, each defensible on its own terms, each reinforcing the illusion that the business was scaling naturally.
One fact that appears repeatedly in the record is how quickly scale itself became part of the seduction. Global Crossing’s network was vast enough that outsiders could mistake sheer geographic reach for financial solidity. A company with cables in oceans and switching facilities in major markets looked too large to be built on illusion. Yet the evidence later examined in securities litigation and accounting review suggested that the company’s reported momentum depended on arrangements that were more circular than organic. When the market rewarded expansion, the pressure was to keep the expansion narrative alive.
The first money flowed through ordinary corporate channels: capacity leases, interconnection arrangements, swap deals, and revenue recognition entries that treated exchange as sale. In a market obsessed with growth charts, the figures on the income statement became a second network — invisible, but more convincing than any map. Once that machinery was running, it no longer needed to shout. It just had to keep moving.
And that is what made the setup so dangerous. If the revenues were inflated by transactions that merely rotated capacity among participants, then the company’s reported health could remain strong long after underlying demand failed to justify it. Investors would see continued growth; lenders would see continued scale; analysts would see continued momentum. The hidden risk was that the company could build its reputation, its valuation, and its financing structure on a foundation that was less a market than a loop.
The tension inside that loop was practical as well as moral. The larger the company became, the harder it was for outsiders to separate legitimate infrastructure buildout from accounting that made growth appear cleaner than it was. The more the market trusted the story, the less likely anyone was to ask why demand was not translating into corresponding cash. That meant the earliest warning signs were easy to miss, especially when industry-wide enthusiasm made caution look out of step with the moment.
And once the books had begun to move on their own, the problem was no longer whether the company could sell bandwidth. It was whether anyone inside the building still knew where the real demand ended and the accounting began.