On Air Now

Upcoming Shows

Program Schedule »

Listen

Listen Live Now » 1450 AM Holland, MI

Weather

Current Conditions(Holland,MI 49422)

More Weather »
70° Feels Like: 70°
Wind: E 5 mph Past 24 hrs - Precip: 0”
Current Radar for Zip

Today

Partly Cloudy 85°

Tonight

Partly Cloudy 67°

Tomorrow

Partly Cloudy 86°

Alerts

Google-based trading strategy would have made big return: study

The Google signage is seen at the company's headquarters in New York January 8, 2013. REUTERS/Andrew Kelly
The Google signage is seen at the company's headquarters in New York January 8, 2013. REUTERS/Andrew Kelly

By Sudip Kar-Gupta and Alistair Smout

LONDON (Reuters) - A trading strategy based on changes in the number of times financial phrases are searched on Google could have generated returns in excess of 300 percent over the last decade on the Dow Jones, a study has found.

The findings could prove particularly useful for computer-driven trading firms that use the Internet and Twitter to help gauge investor sentiment.

The study by academics from Warwick Business School, University College London and Boston University, analyzed changes in the frequency with which 98 terms such as "revenue" and "unemployment" were used in searches from 2004 to 2011.

Using the data to trade the Dow Jones Industrial Average Index would have returned a "substantial profit", they said.

The study examined volumes of searches for a list of keywords - which Google makes publicly available - and sold the Dow Jones short at the beginning of the week if search volumes on the word were higher than the previous week, and bought the index if volumes were lower.

"The initial hypothesis was that an increase in search volume was a sign of investor concern," said co-author of the report Dr Tobias Preis of Warwick Business School.

By employing the strategy on the keyword "debt", for instance, an investor would have seen a return of 326 percent since 2004.

(For the full report please click on: http://www.nature.com/srep/2013/130425/srep01684/full/srep01684.html)

(Editing by Simon Jessop and Mark Potter)

Comments