Stock Futures

Innovation by the Book

Putnam is a Wall Street maverick. But he's no reckless rule breaker. A favorite tactic is to learn the rules of Wall Street intimately and use them to his advantage. That's how he launched Archipelago.

In 1996, trying to make a go of his business, Putnam read a 1,000-page tome that spelled out new order-handling rules from the SEC. The rule set was handed down after a two-year investigation into claims that Nasdaq dealers kept the spreads between stock prices artificially high. One outcome of the SEC edict was that it allowed electronic networks to link directly to Nasdaq and let market makers access limit-order books through Nasdaq systems. Putnam realized what that meant: An electronic network could fully integrate its systems into Nasdaq, display its quotes, and have those quotes accessed. The ruling could make his outfit a player.

The SEC knew that the three existing, private electronic-trading systems -- Bloomberg Tradebook, Instinet, and Island -- would apply to become ECNs. "But the SEC never expected to hear from a startup in Chicago," says Putnam. "I told them we'd be ready when the rules went into effect in January 1997. I don't think they realized that they had created a new industry."

Putnam and his team had just eight weeks to nail the deadline. He suspected that regulators would approve just one new company and then wait until 1998 before permitting other upstarts to follow. That would give him a 12-month jump on the inevitable competition. Fortunately, he already knew two people who could build an ECN in record time: Stuart and MarrGwen Townsend.

It was Stuart Townsend who gave the new network the name Archipelago, a clunky word that sums up a radically new model for executing trades. The SEC's biggest objection to Putnam's application was this: Unlike Island and the other networks, Archipelago didn't have access to a deep pool of liquidity. Putnam said he didn't need one. His plan was to leverage an open-architecture approach that would link his system not just to one single island of trading but to a group of islands -- an "archipelago" -- of networks and stock exchanges. If Archipelago couldn't find the best price on its own order book, it would find it somewhere else, whether that meant going to Nasdaq, the NYSE, the American Stock Exchange (Amex), another market maker, or even another electronic network.

Archipelago's open-architecture model was a triumph for the democratization of information through technology. It and the other electronic networks cut out dealers and other middlemen, allowing institutional buyers and sellers to post prices directly to each other. As a result, transaction costs tend to be lower than on the exchanges, and executions are faster. Trades on Archipelago's system are completed in less than a second, while a typical man-made trade on the Exchange can take 16 seconds. "That's a lifetime in a volatile market," says Putnam.

Putnam is pushing speed and transparency to another level on ArcaEx, the first U.S. stock exchange to publish an open book that displays every pending limit order. Unlike traditional exchanges, there are no insiders with information advantages on ArcaEx, because prices and share quantities are broadcast to the trading world in real time. Bottom line: There are no insiders with trading advantages. "There's a reason why it costs millions for a seat on the New York Stock Exchange," says Putnam. "A seat on the Archipelago Exchange costs nothing, because it's worth nothing. Everything is transparent."

Unfinished Business

On an early morning in March, Putnam and some other executives gathered around a PC and watched as the Archipelago Exchange executed its first trade, on a Pacific Exchange listing called the Thai Fund. Not exactly a heavy-volume security, but that's the point. The plan is to start small, keep testing the system, and gradually phase in Amex, Nasdaq, and NYSE stocks.

As he watched ArcaEx take its first tentative steps, Putnam says that it felt a little anticlimactic. For a moment, he thought about everything that Archipelago had accomplished since 1996: It had built the exchange. It had the jump on technology. And in those server farms, it had enormous bandwidth. But in a sense, it had taken six years just to get to the starting line of what's shaping up to be a marathon race for market share.

The NYSE has launched a counterattack with Direct+, proof that the success of the electronic networks has pushed the Big Board to innovate. "The Exchange has huge credibility," says Adam Townsend, a J.P. Morgan Chase analyst. "If the Exchange can get things right on the technology side, traders will continue to work with it directly. If it can't, the orders will migrate to the most efficient platform."