The Gateway to Algorithmic and Automated Trading

High-Frequency Trading: Alpha Discovery and the New Arms Race

Xand : pcurley - 1st January 1970

The opinions expressed by this blogger and those providing comments are theirs alone, this does not reflect the opinion of Automated Trader or any employee thereof. Automated Trader is not responsible for the accuracy of any of the information supplied by this article.

high frequency trading arms racePerhaps nowhere more in the diverse world of hedge fund strategies is the prospect of alpha decay more unsettling than at high-frequency trading firms. In many ways high-frequency trading firms are now facing a reality that other hedge funds with more esoteric strategies may one day face - too much money chasing a finite amount of alpha.

The End of the Hardware Arms Race

The latest figures indicate that high-frequency trading now accounts for somewhere between 60-70% of trading volume in the US. This is up from around 35% just five short years ago. High-frequency trading firms, that in the past have been among the most profitable on Wall Street, are now seeing that increased competition has crowded out many of their traditional strategies. High-frequency trading firms have responded by co-locating their black boxes and by throwing ever more expensive hardware at the problem. This approach has worked for some of the larger, better-funded firms, but only postpones the inevitable.

What will happen when this hardware arms race meets the laws of physics? The answer is that many of these firms will need to develop high-frequency trading strategies where alpha is still relatively abundant and competition less fierce.

high frequency trading market data cloud

New High-Frequency Trading Alpha Opportunities

The two most common opportunities that high-frequency trading are now exploring are as follows:

  1. Alpha outside the US - The US was the first to develop electronic trading, other regions are still in the process of building out their electronic trading infrastructure. A recent Credit Suisse study estimated that high-frequency trading activity accounts for 35% in Europe and only 10% in Asia (excluding Japan). Many high-frequency trading firms are now seeking to port the strategies that worked so well in the US over to these less developed markets.
  2. Multi-Dimensional Strategies - The traditional strategies of high-frequency trading have typically been one-dimensional involving the high-frequency trading in and out of a single large, liquid name. The newer high-frequency trading strategies are much more complex and multi-dimensional in nature searching for arbitrage opportunities across asset-classes, geographies etc.

The New Arms Race and Cloud-based Market Data

The common thread that runs through these new strategies is the need for access to more diverse and dispersed market data sets. Market data has always been the lifeblood of high-frequency trading firms. It has also been one of the real pain points of this industry with the requirement to build out large infrastructures to manage and query terabytes of information. This new world of alpha discovery means that high-frequency trading must now potentially analyze vastly greater amounts of data to discover opportunities. Furthermore, the competition to discover these new opportunities will be intense. In this new world alpha will also be more fleeting as firms quickly clone successful strategies. The new high-frequency trading arms race will be all about alpha discovery and time to implementation.

As the high-frequency trading industry's search for alpha switches from a hardware focus, to a discovery/speed of implementation focus, it is clear that firms with access to large and diverse data sets, that are easy to query and work with, will be at a distinct competitive advantage. The traditional method of downloading vast data sets to a firm's own data center will not work because of the cost and time delays involved. The only solution that is flexible enough to assist in such rapid discovery and implementation is a Cloud-based approach, where small sub-sets of global multi-asset class market data can be quickly isolated and analyzed. Xignite is the only firm committed to this vision of large, diverse sets of Cloud-based data, all made accessible by our easy-to-use web services, that require no infrastructure, and that can implemented in a matter of minutes.

As this new arms race intensifies and the search for alpha becomes increasingly more desperate we will see more and more high-frequency trading firms abandon their in-house data management infrastructures for the flexibility and breath of data that Xignite's Cloud-based solution offers.

Share and Enjoy: email Twitter Facebook LinkedIn Digg StumbleUpon Technorati Suggest to Techmeme via Twitter Google Bookmarks Live FriendFeed