Sociotechnological implications: Levelling the Stock Market Playing Field

By | April 2, 2014
Sociotechnological implications: Levelling the Stock Market Playing Field
Unifying the very small and the very large

Ever wonder how high-speed trading gets people rich? It's a very simple algorithm, really, and can be beat by an ingeniously simple solution:
1) Watch for a 'slow' buy order coming from a real person
2) Issue the same order at another exchange before the first trade is executed
3) Now you've driven the price up a few cents, so wait a few milliseconds until the 'slow' trade is filled, and then issue a sell order.

The advantage lies in the network fibres: The algorithm assumes that their high-speed fibre will get the order to another exchange before your order gets there on the 'slow' cables.

Now the solution: send the same order with a slight time delay to all the exchanges, so that they arrive at the same time. This prevents high-speed traders using that particular algorithm from gaming the market. That's some deep-level packet and hardware knowledge needed to beat the market.

Which raises an interesting insight that applies to not just stock markets but also to other aspects of daily life, as ubiquitous computing brings lower-level interfaces into consumer gadgets: Technology changes our habits and processes at a fundamental level. To understand this change, we need to understand how these fundamental, low-level hardware concepts create implications on the higher level of our society.

Translated into English: The 1% understand that the nitty gritty details which network geeks deal with can line their pockets with gold, because they assume that no other sane person would care to look at both the nitty gritty details and the big picture.

#Society   #Technology   #Ethics  

/via +Ansel Taft 

Is the U.S. stock market rigged?
Steve Kroft reports on a new book from Michael Lewis that reveals how some high-speed traders work the stock market to their advantage

7 thoughts on “Sociotechnological implications: Levelling the Stock Market Playing Field

  1. Sophie Wrobel

    +Adam Hartman that's true, there's always some randomness in everything. But the order fill rate changing from never to 100% is telling. That said, given the amount of money involved, I expect people will try to game the new game.

  2. Sophie Wrobel

    +Adam Hartman it's based on the length of the cables running between the exchanges, and the type of cables they are. Signals move only so fast through a given material; that's why the network engineers who laid the cable were called in to provide the details.

  3. Adam Hartman

    How do you know what length of delay to add to the stock signal?  That's the problem.  Calculating it adds more delay, making it harder to calculate, adding more delay.  You can guess, but that won't make them "arrive at the same time".  Some will still win, and can continue to win if they can reverse engineer the way the delay was calculated/guessed.

  4. Dirk Reul

    This is why proximity to the Exchange's trading hubs is so important and low latency connections are something worth a lot.

    Btw, this can also be done with crypto currencies like Bitcoins and Dogecoins etc, by buying and selling and buying the same set of "coins" back and forth between different exchanges, you a) make sure the prices are averaged out over different exchanges b) you make money on the different current prices for the crypto currency. This only works of course if you have a scripting engine / auto trading system running that compares these continuously.  While this will of course generate nothing like what happens in high speed auto trading, it is nonetheless quite fascinating to see develop in alternative systems as well. 


Leave a Reply

Your email address will not be published.