Search

Listening to the Dow

U-M librarian transforms stock exchange ups and downs into sound waves

June, 2011

By Lynne Raughley

Justin Joque, Spatial and Numeric Data Services Librarian at the University of Michigan Library, frequently works with U-M researchers in various fields to collect and render data in ways that illuminate their studies. Data visualization—that is, transforming numeric information into graphical form—can make complex information more readily understandable, and can lead to new insights and knowledge.

Joque notes that while scientists are increasingly interested in mapping data onto other media in the hope of revealing patterns that are otherwise invisible, artists are increasingly seeking out data sources, either for material or inspiration.

Recently, he became interested in data sonification—that is, rendering numeric information into sound waves—and wondered what it might be capable of revealing, as both research and art.

With no musical background, Joque decided to keep it simple. He used an open source programming language and environment called Processing, and created two sine waves that range from 20-800 herz. One of the waves maps to the daily changes in the Dow Jones Industrial Average, and the other to the daily number of trades, from the late 1920s to early 2011 (data he obtained from Yahoo).

“The period after the Great Depression through the 1960s is somewhat tedious,” Joque says. “But I think it’s worth listening to in order to hear the advent of high frequency trading.”

Asked when that happened, Joque smiles. “Just listen,” he says. “You’ll hear it.

10 Comments

You can follow any follow up comments to this entry through the RSS 2.0 feed.

10 Comments to Listening to the Dow

  1. by Michael Flynn

    On June 17, 2011 at 10:30 am

    I love this idea…..where’s the link to the sound wave? Why did Justin Joque map the data into a sine wave? Would the unaltered data waveform sound too noisy?

  2. by Margaret Hedstrom

    On June 30, 2011 at 7:36 am

    Way to go Justin. I know that data visualization is a big deal, but I love data sonification.

  3. by Ian Demsky

    On June 30, 2011 at 7:51 am

    Very interesting. I’d be curious to know if inflation was factored in.

  4. by Rainey Lamey

    On June 30, 2011 at 8:49 am

    This was brilliant, Mr./Dr. Joque! It was remarkable to hear the original stock market crash and JFK’s assassination “bloop”s, as well as the too greased frenzy of the 90s market. To be able to see and hear “the wheels come off,” so to speak, was simply jaw-dropping. Thank you!!

  5. by Tyler Frankenstein

    On June 30, 2011 at 9:31 am

    Sounds like we’re in for an interesting ride ahead.

  6. by Sabrina

    On June 30, 2011 at 11:07 am

    It’s a strange sensation when, after 6 minutes and 1 second, all goes quiet.

  7. by Linda Knox

    On June 30, 2011 at 11:18 am

    So fascinating and profound. Is there an emerging community of interest in data sonification at UM? Very cool to have a Librarian’s initiative out front.

  8. by Justin Joque

    On June 30, 2011 at 1:24 pm

    Thanks everyone for the comments and feedback!

    Michael Flynn – There isn’t a copy of just the sound posted, but the video should play it. The main reason I used a sine wave was so I could easily combine the trading volume with the daily change. Also, the way the software I used was set up this seemed much easier for an initial exploration, but I am planning on trying some different things with this and other data in the near future – so I will definitely give it a try.

    Ian Demsky – The data doesn’t account for inflation, but I used the daily change in closing prices (in addition to trading volume). So since each day is only compared to the day before the effect of inflation should be pretty minimal. On the other hand the trading volume data is just the raw volume, hence the intensification after computer trading starts.

  9. by Anthony Oliver

    On June 30, 2011 at 2:22 pm

    Very cool indeed. Michael, my guess is the sine wave probably helps normalize the data, then outliers probably cause bigger pitch changes from the “norm”.

    I’m also amazed how many people have never heard of HFT.

  10. by William H Murphy

    On September 29, 2011 at 7:11 pm

    It is interesting that you used a sine wave while woking with DJIA data. In 1970 an electrical engineer named J M Hurst ” The Profit Magic of Stock Transaction Timing”,figured out the complete mathematical and conceptual explanation to how all markets work by using Fourier analsis and a comb of digital filters. It turns out that the basic building block of all stocks and commodities is the sine wave. Markets are simply a sumation wave of a certain set of harmonically related sine waves which vary in period, magnitude and phase. If Justin Joque is interested in seeind the proof then contact me at whmurphy2@yahoo.com.

Comment

By submitting a comment here you grant Montage a perpetual license to reproduce your words and name/web site in attribution. Inappropriate or irrelevant comments will be removed at an admin's discretion.