Over a million developers have joined DZone.

The micro market crash, brought to you by social media

DZone's Guide to

The micro market crash, brought to you by social media

· Big Data Zone ·
Free Resource

The open source HPCC Systems platform is a proven, easy to use solution for managing data at scale. Visit our Easy Guide to learn more about this completely free platform, test drive some code in the online Playground, and get started today.

Just over a week ago, the U.S. stock market plunged in early afternoon trading with the S&P 500 losing $121 billion of its value within minutes. This happened after a fake tweet appeared on the Associated Press’s twitter feed describing a terror attack on the White House.  During the momentary plunge, the Dow fell 145 points or roughly 1%, turning briefly negative for the day after being up more than 135 points. It rebounded as quickly as it fell.

Many point to computerized trading algorithms as the trigger for the rapid decline and subsequent rebound.  Algorithmic trading is the use of high speed computers for entering trading orders with an algorithm which executes pre-programmed trading instructions based on certain market conditions; in this case the algorithms were ‘faked out’ by the fake tweet that appeared on the Associated Press’s twitter feed.

With the Quick Rebound, Where is The Harm

If the market rebounded quickly after the “flash crash”, what’s the harm?  Well, not everyone was able to sell and buy at the exact time to ensure they were not affected by the crash.  There were most certainly big winners and big losers as a result of this flash crash. The faster the market gets, and algorithmic trading is hyper-fast, the more there are have’s and have not’s.

What Separates the Winners from the Pack?

To understand who wins and who loses in these types of events you need to understand the very basis of how financial exchanges operate.  Exchanges maintain an order book that keeps track of a list of the entire buy and sell orders for the exchange. They use software called a ‘matching engine’ to match the buy and sell orders that best correspond to one another.

Now consider the algorithmic trading systems that automatically trigger off of different events in the market.  If you can be just a little faster than the other algorithmic trading systems your orders will always be the first to be executed at a given price, or in a down turn your orders can be canceled just a little faster than your competitors.  It was not so long ago that speed was measured in milliseconds but the investment firms have been in an arms race-type battle over who can create a system with the lowest latency allowing them to trade faster than the competition gaining the competitive advantage.

Today the competitive advantage is coming down to microseconds.  A microsecond is equal to one millionth of a second and every little thing counts when you are looking for a competitive advantage of a couple of microseconds.  To amplify this point, broker dealers who were not located as close as others to Exchange computers could not get their orders in as quickly even though they were using fiber optics and communicating at the speed of light.  As a result Exchanges are now offering colocation services where the broker dealers obtain rack space for their servers within the same datacenter as the Exchanges maintain their systems. It’s hard to get faster than that.

The Next Battleground

Middleware vendors are beginning to leverage specialized hardware that allows for the deploying of applications directly onto the network layer via an optimized Server Switch, eliminating multiple servers and removing additional network hops to further reduce end-to-end latency.  These specialized network switches are fully integrated solutions that combine Layer 2/Layer 3 switching fabric that natively integrates with existing networking backbones and the ultra-low latent messaging service platform.

For the early adopters, they will be able to more quickly respond to market events to better serve their customers. For those who choose not to keep up…well…how long before the next flash crash?

Managing data at scale doesn’t have to be hard. Find out how the completely free, open source HPCC Systems platform makes it easier to update, easier to program, easier to integrate data, and easier to manage clusters. Download and get started today.


Opinions expressed by DZone contributors are their own.

{{ parent.title || parent.header.title}}

{{ parent.tldr }}

{{ parent.urlSource.name }}