The top three competitors in this year’s Stanford Women in Business (SWIB) Fantasy Stock Exchange (FSE) used a trading algorithm for the first time in the competition’s history, allowing them to make thousands of trades and achieve daily rates of return above 20 percent.
The algorithm, which was created by Andrew Han ’16, Jessica Xu ’16 and Miraj Rahematpura ’16 during spring break, was applied towards the end of the 10-week competition. Han, Xu and Rahematpura were propelled to the top of the leaderboard, displacing competitors who had been in the lead for weeks.
Each of the 135 competitors was initially given $1,000,000 to make virtual trades. Han placed first with total returns of $509,750.42, winning a $300 prize from SWIB. Rahematpura came in second and won a $200 prize for his returns of $421,925.43, and Xu placed third with $397,832.56 in total returns, winning $100. The fourth place competitor, Lucas Thompson ’16, had total returns of $240,398.16.
According to Han, the algorithm is programmed to purchase high volumes of cheap and volatile stocks, such as Zynga and Groupon, wait for small gains in the price and sell the stocks to profit off these price increases.
“We found stocks that were really volatile so they were constantly changing, and since the computer could constantly update it while we couldn’t ourselves, that’s why our stocks went up a lot more than if we did it by hand,” Xu explained.
Han, Xu and Rahematpura each played a different role in the algorithm’s development, with Xu researching stocks, Han writing code and Rahematpura doing the mathematical calculations. While the three used similar algorithms, they invested in different stocks and made different numbers of trades.
“The algorithm would look at stock prices, and if certain conditions held, it would buy. If certain conditions held, it would sell,” Han said. “It was sort of a learning experience for us, too. [Xu] selected a couple of stocks and we would each choose different ones that would interface with whatever combination of algorithms that we had set.”
Though Han said that he would understand if other competitors were “a little miffed” by the use of the algorithm, Thompson, who had been in the lead for four weeks before the last week of the game, thought the algorithm was fair.
“I think if you can come up with a trading algorithm that works that well, that’s awesome,” Thompson said. “Clearly it worked extremely well. If you are willing to put that amount of effort into it to come up with something like that, I have no problem with that.”
Sohaib Shaikh’16, who came in sixth in the competition, said that creating algorithms is “really what a lot of the stock market has become nowadays.”
“It is honestly about who can develop the fastest algorithm,” he said. “In the game, sure, I think it’s a completely reasonable thing for people to use [algorithms]. It is an opportunity for them to test these algorithms.”
According to Paige Gonye ’14, director of FSE, the use of a trading algorithm in the FSE is “unprecedented.” After speaking with Han, Xu and Rahematpura to ensure that they had each contributed to the algorithm, Gonye determined that the use of the algorithm, though unconventional, was fair.
“We were definitely surprised, but true to Stanford fashion there is always a new approach that students are willing to experiment with,” Gonye said. “Because they were able to prove that they each contributed something to the project and they each have said exactly what they contributed, they are entitled to the prizes.”
While Gonye said that Han, Xu and Rahematpura earned their prizes, she noted that SWIB will likely restructure the game in future years, either limiting the number of trades or creating two games—one where the use of algorithms is encouraged and one where a long-term strategy is emphasized.
“The game has continued to change and evolve over the years, but this is going to probably be a larger change than we would have made,” Gonye said. “We are trying to teach students who aren’t as sophisticated necessarily, and we want to make the game fair.”
Han said that he would participate in the game next year if SWIB decides to offer a second game that encourages the use of algorithms. However, he said that that he would be wary of using trading algorithms again in the normal competition.
“It’s kind of like a weapons of mass destruction kind of thing, one or two weeks of running the program and there’s no point in even trying anymore,” Han said. “To preserve [SWIB’s] purpose, which is to teach people the basics of investing, I wouldn’t join it.”