The algorithm also accounted for a mammoth ten percent of the bandwidth that is allowed for trading on any given day, but the motive behind its presence is still uncertain.
It placed orders in 25-millisecond bursts, which included about 500 stocks, but never actually executed a single trade. The algorithm's activity ended at 10.30am ET on Friday.
"Just goes to show how just one person can have such an outsized impact on the market," Eric Hunsade, head of Nanex, told CNBC.He added: "Exchanges are just not monitoring it."
Although the exact motive behind the algorithm is still unclear, it that a trader could have been testing the water to see how he or she could use the additional bandwidth to create a latency advantage.
By using up additional bandwidth, the program could potentially slow down others trying to execute trades, allowing it to buy and sell on information that its competitors wouldn't receive until fractions of a second later.
High frequency and electronic trading is proving to be a major problem for regulators that want to better control the practice, which accounts for 70 percent of trading each day, but without restricting profits in the City and on Wall Street.
Hunsader told CNBC that regulators should do something fast to better control such situations, as this type of algorithm could have had a negative impact on the stock markets if big news had broken during the week.
It has been a turbulent year so far for technology in the financial markets, where a number of firms and exchanges have suffered technical glitches.
Most notably that caused losses of about £281 million and during the much anticipated Facebook IPO, which resulted in market making firms incurring hundreds of millions of dollars in losses.