Nearly 30 years ago, Traders Magazine took a look at a new kind of technology beginning to make its way onto Wall Street trading desks.

In a June 1998 article, “For Your Eyes Only: In Undercover World of Intelligence, Agency Desks Utilize New Technology”, the magazine wrote about so-called “intelligent trading systems” and the mathematicians and scientists behind them. Among them was Dr. Mark Gimple of Reynders, Gray & Co., who was affectionately nicknamed “the walking algorithm” by his colleagues.

Gimple had developed a system designed to lessen market impact by tracking daily trading volume in NYSE-listed stocks and determining when volume was likely to peak. Once a trader entered an order’s price range and time frame, the system could decide when to execute it.

The system cost more than $400,000 to develop and took six months to get running. It could handle as many as 6,000 client orders a day, although Nasdaq orders still required more human involvement. By 1998, almost half of Reynders Gray’s average daily volume of 1.5 million shares was already being executed through the system.

Gimple believed this was where agency trading was headed. He told Traders Magazine that investors were increasingly looking worldwide for liquidity and would need technology to help find it.

Later in the article, he returned to the same point:

“Investors are looking worldwide for liquidity, and they increasingly need technology to find it. … You need the smartest technology to stay ahead.”

Nearly three decades later, that logic sounds considerably less futuristic. Electronic trading systems routinely search for liquidity, assess market conditions and help determine how and when orders should be executed.

What stood out as an “intelligent” trading system in 1998 has become part of the everyday machinery of modern markets. The technology has changed dramatically, but the goal remains finding liquidity and executing orders efficiently while minimizing market impact.

In 1998, those systems were still novel enough that the scientist developing one was known around the office as “the walking algorithm.” Today, the ideas behind his work hardly sound unusual.