July 27, 2009
A controversy is brewing regarding one of the most lucrative HPC applications ever invented: high frequency trading. Also know as algorithmic trading, high frequency trading (HFT) is the process by which computers are used to execute trading orders at extremely low latencies -- on the order of milliseconds. The speed of the HPC machines and the associated communication technologies makes this all possible.
According to the TABB Group, HFT accounts for 73 percent of all equity trading volume in the US and those trades are executed by only about 2 percent of the trading firms operating in the country. At more than $21 billion annually, HFT-generated revenue is on par with that of commercial gambling.
In fact, in some ways HFT operates like a low-latency casino, and not a particularly honest one at that. In a recent New York Times article, it was pointed out that HFT players are using special access to data streams and high powered technology as a way to game the system. Retail and institutional investors without this type of setup are at disadvantage. From the article:
High-frequency traders often confound other investors by issuing and then canceling orders almost simultaneously. Loopholes in market rules give high-speed investors an early glance at how others are trading. And their computers can essentially bully slower investors into giving up profits — and then disappear before anyone even knows they were there.
The "loopholes" are something called flash orders, whereby bids that are not immediately filled are "flashed" to a proprietary data feed that participants can buy into. Once there, the computers can do their work out of the public eye, taking advantage of the private information to know when to buy and sell at the most opportune price points. For example, the algorithm can buy low as it sees demand building, and sell high as demand grows (which it has helped fuel). If mere mortals were doing this, it would be called front running, which happens to be illegal.
Algorithmic-driven transactions that buy low and sell high often yield just pennies per share on individual trades. But over billions of trades this can add up. Another New York Times piece pointed out the major problem with the flash scheme:
Although anyone can gain access to flash orders by paying a fee, they are useful only to traders who have computers powerful enough to act on the data within milliseconds. In recent years, some of the largest financial companies, including Goldman Sachs, have earned enormous profits with such computers, which are very expensive and often housed right next to the machines that power the marketplaces themselves.
In a recent interview (video) on Bloomberg, Former NASDAQ chairman Alfred Berkeley noted that HFT players are not looking for long-term investments, but rather for "temporary imbalances in supply and demand," which can be exploited in very short time frames. At one point, Berkeley characterized the HFT players as "scalpers," yet he maintains that they serve a real purpose in maintaining market liquidity. According to him the real question is if HFT is putting the average investor at a disadvantage. His take is that the current structure is probably skewed toward speculation rather than investing, and needs to be rebalanced.
For more of a point-counterpoint view of the topic, I've posted a recent CNBC segment with Joe Saluzzi of Themis Trading and Irene Altridge Partner at Able Alpha Trading. Pretty interesting exchange.
Posted by Michael Feldman - July 27, 2009 @ 5:20 PM, Pacific Daylight Time
Michael Feldman is the editor of HPCwire.
No Recent Blog Comments
10/30/2013 | Cray, DDN, Mellanox, NetApp, ScaleMP, Supermicro, Xyratex | Creating data is easy… the challenge is getting it to the right place to make use of it. This paper discusses fresh solutions that can directly increase I/O efficiency, and the applications of these solutions to current, and new technology infrastructures.
10/01/2013 | IBM | A new trend is developing in the HPC space that is also affecting enterprise computing productivity with the arrival of “ultra-dense” hyper-scale servers.
Ken Claffey, SVP and General Manager at Xyratex, presents ClusterStor at the Vendor Showdown at ISC13 in Leipzig, Germany.
Join HPCwire Editor Nicole Hemsoth and Dr. David Bader from Georgia Tech as they take center stage on opening night at Atlanta's first Big Data Kick Off Week, filmed in front of a live audience. Nicole and David look at the evolution of HPC, today's big data challenges, discuss real world solutions, and reveal their predictions. Exactly what does the future holds for HPC?