HFT (High Frequency Trading) is the last stage in the evolution of computerized algo trading, where the last frontier and variable to optimize is execution speed. HFT algorithms do not do anything more sophisticated than what pit traders, market makers or even technical traders did in the past. Their only difference is the speed by which they accomplish their objectives. The fact that there are some sophisticated HFT algorithms cannot support a claim of a new and revolutionary approach to trading.
The advantage of running HFT algorithms is that at this stage of the game most retail traders cannot afford their implementation. Thus, these ultra-fast trading systems have an advantage over those who transmit their orders at slower speeds. However, if one could imagine a market where only HFT algorithms compete, then the high profitability of those systems will again be forced to conform to some kind of distribution where 10% -to 20% of the participants share 80% to 90% of the profits (Pareto type distribution). Thus, as long as there are slow intraday and short-term traders, HFT systems can maintain an advantage and be profitable as a group but they are not anything new in terms of a trading concept. They are basically ultra fast algorithmic trading systems that are based on concepts developed during the past three decades.
A New Generation of Trading Systems
A completely new generation of trading systems will emerge in the next 5 to 10 years and will not depend on anything that most are familiar with nowadays, i.e. order book arbitrage, tape reading, triangular forex arbitrage, event arbitrage, you name it. These new trading systems will be based on a totally different theory and will establish a new paradigm. There are scattered reports of such systems being developed and I have also been designing some on paper. The fact is that trading systems will be much different from what they are nowadays ten years from now but this field will be much more secretive. In addition, the new generation systems will be high-priced because they will offer an edge .
Let us talk about price because money has lost its value and even meaning lately. When Omega Research Inc., nowadays known as Tradestation Technologies Inc., developed the first efficient backtesting software program known as System Writer Plus, back in the mid 1980s, it was priced at $2,000 a copy. Using that program at that time offered an edge over others who were not familiar with backtesting and my trading partner and I bought not one but two copies for $4,000. Using that tool we were able to backtest an idea that helped us grow a $200,000 account to about $350,000 in a matter of a few months, trading currency futures. Although that software had several constraints, we were able to profit from it before it became a widespread accessory and it along with naive backtesting lost any edge it could offer.
Why is it then that nowadays traders think that spending $1,000 or even $500 for a trading program is too much and are even looking for free tools which exist in abundance? That is mainly because of the high competition in this area that has basically driven profit margins down to zero. The new generation of trading systems to appear in the next 5 to 10 years will again create value for an industry that is barely surviving due to proliferation and provide an edge to those who will be able to afford them.