There have been some positive developments in the past few weeks, but since I spent my Spring Break in Austin I did not get as much work done on my thesis as I hoped to.
Access to Intraday Trading Data
I am now in touch with Knight Capital to get one day’s worth of exchange data. When I met with Marius Watz for my 1-on-1 meeting a few weeks back, he mentioned that he’d done a project for them, called Stockspace. He told me to drop his name and see if they’d share data with me like they did for him. Finally, after a few weeks of emails, they are working to get me data for a day’s worth of exchange activity, as well as one day’s worth of Knight trading activity. This will really help me get a feeling for the intraday behavior of the trading exchanges, which is especially important since high frequency traders buy and sell during the day, but try to leave their position “flat” at closing. Basically, this means they don’t want to end the day holding shares of stock whose values might drop. I hope to use Mark Hansen’s Data class to really explore the intraday data and see what kinds of patterns and visualizations emerge.
I may use this data to animate a similar landscape in Unity like I’ve been experimenting with this semester. Here is a video of the latest version, though still not where I envision it in the long run.
Hunting down data for Colocation Map
I spent a lot of my Spring Break hunting down data on the locations of trading exchanges, as well as colocation facilities that service the proprietary trading firms that engage in high frequency trading. Following Dave Boyhan’s suggestion, I did a lot of whois.com searches on the IP address from this not completely inclusive list of these prop trading firms. It returns Latitude and Longitude coordinates, but I’m not quite sure if they’re really accurate. I’m also torn about whether to list the trading firms at their official addresses, or to track down where they colocate. I will try and do both, but again, there’s a lot of private data so I’m also limited to what I have access to.
Spread Networks, which is currently the fastest fiber cable between New York and Chicago, making a roundtrip in 13.10 ms, updated their website recently and made a lot more of their locations public. They also posted a map of “amplification sites” along the way. This makes me really want to roadtrip to all of these towns, but I’m afraid that dream may have passed since I spent my Spring Break eating tacos in Austin.
Still, it’s great that they’re publishing more of their data. I’d like to contrast the Spread Networks cable with a few that are a tiny bit slower, but are much cheaper because of it. I’m assuming that the real HFTs use Spread Networks since even 1ms advantage is worth the cost in most cases.
In addition to the Klondyke Gold Rush map, I’m looking at some maps Stamen made of London, in which they visualized the relationship where they made a heat map that corresponded to the time it took to commute downtown.
I’ve spent a while weighing the pros and cons of this since I last presented it in Thesis class a few weeks ago. I’m still fascinated by the extreme importance of time keeping in HFT, and the fact that it needs to be accurate down to the microsecond. I have ideas like making a Time Microscope or a High Frequency Time Machine, which would be imaginative objects that address the conflict of humans trying to experience computer time, or the lengths that traders might go to find that edge.
When I talked about the GPS Spoofer in Crit Group, I received interesting feedback that made me question why I wanted to make it. Abigail Simon said, the financial system is vulnerable (and problematic) due to larger issues than GPS timekeeping. I’m torn between exploring this technical facet vs. pushing some more conceptual ideas I have, especially about the relationship between finance and landscape/geography.
I’m still going to try and make the spoofer, and will hopefully have a report on that soon. I’m just not sure how to frame it. Is it a “guerilla weapon?” A sculpture? Do I have to test it in order for it to be worth doing?
Research Group Paper
I’m going to take the paper I’m writing in Heather Dewey-Hagborg’s Research Studio as an opportunity to write what I’ve been thinking about High Frequency Trading down in some kind of more formal way. I will write an update with more specific details, but I see this paper as both explaining High Frequency Trading to a general audience, as well as writing about the issues of time and space, geography and value that my thesis is exploring.