Algorithmic trading (AT) has been one of the most discussed topics in the
financial industry. Algorithmic techniques and the technology that powers them
are now highly influential in the way that financial instruments are traded.
Once the exclusive domain of investment banks, prop trading firms and hedge
funds, AT is in the process of going mainstream.
AT was initially used exclusively in equities markets, however its use has
spread to other asset classes, including futures, options, and most recently
foreign exchange (FX). A third of all stock trades in 2006 were driven by
algorithms. As of 2009, algo trading firms accounted for 73% of all US equity
trading volume. The TABB Group has estimated that the total revenues from such
trading amounts to $21 billion annually.
Originally, the term algorithmic trading was used to describe any type of
computer assisted trading activity, which handles the timing, submission and
management of orders. However, as AT techniques evolved, and new methods of
utilizing AT emerged, the term has expanded to encompass other terms, such as "program
trading", "auto trading", "black box trading" and
"high frequency trading" across single or multiple pools of liquidity.
The major reasons for the creation of AT and the trends driving the growth of
algorithmic trading are an increased uptake of electronic trading methods,
decreasing margins, speed of markets, combined with a multiplicity of execution
venues, and a focus on cross-asset class opportunities.
ECNs and fully electronic execution were developed in the late 80's and 90's,
and in 2000 equities were "decimalized" and the minimum tick size was
changed from 1/16th of a dollar ($0.0625) to $0.01 per share. This may have
given birth to algorithmic trading as it changed the markets, by narrowing the
spread between the bid and offer prices, decreasing the market-makers' trading
advantage, thus decreasing market liquidity.
This
decreased liquidity led to institutional traders splitting up orders according
to computer algorithms in order to execute their orders at a better average
price. These average price benchmarks are measured and calculated by computers
by applying the volume weighted average price (VWAP). As more electronic
markets opened, other algorithmic trading strategies were introduced. These
strategies are more easily implemented by computers because machines can react
more rapidly to temporary mis-pricings and shop and execute in several markets
simultaneously.
As with any other trading decision, the trader has to determine when to trade,
and how to trade. Algorithmic trading is comprised of these same two elements,
the pre-trade strategy and analysis, and the execution phase. The decision of
when to trade is based on continuously recalculated analytics and monitored thresholds. Once an opportunity is
identified by the pre-trade analysis, the order execution element of the AT
strategy, may slice the order up into smaller orders and place them in multiple
liquidity pools, to insure the best execution possible. Algorithmic trading may
be used in any investment strategy, including market-making, inter and intra
market spreading, arbitrage, or pure speculation, including directional trading
and relative value trading.
As you can well imagine then, an AT solution is a multi-disciplined technique,
where a lot of dimensions come into play when one is being developed. The
primary advantage offered by algorithmic trading is the raw speed of execution.
The ability to pin point and rapidly execute trades has seen a speed "arms
race" between various participants as to who can develop the quickest
algorithm. Large capital expenditures in IT infrastructure, tailored
specifically to the needs of algorithmic trading solutions, have been made to
insure the lowest latency and maximum throughput, thereby guaranteeing firms
will be able to react to market events faster than the competition.
As well as ensuring reduced latency, another criterion for an algorithmic tool
is fitting the right model to the right market. The flexibility and ease of modification that
a particular model allows is important as it means they can be easily adapted and applied to
new market conditions and different asset classes.
In Part 3, I will give the reader a better understanding of how algorithmic
trading works and what it takes to apply this approach to your trading endeavors.
-----------------------------
Gary Phillips is a contributing columnist to www.TheSmallCapInvestor.com. Mr. Phillips is the Founder of GAP
Capital and is a 30-year veteran in trading equities, bonds,
commodities and currencies.
