• 
• 
• 

• Performance
• Services
• Company
• Learn About
• My Attain

System Spotlight: AG Xtreme

July 12, 2004

 

Where have all the good day trading systems gone? The days of LTS Breakout, Dallas Day Trader, and APOSL are long gone. It seems there's just a handful of good day traders left, with very few new ones coming out. Into this environment enters a new entry in teh day trading space - AG Xtreme:

Who is the Developer?

The developer of AG Xtreme is Andrew Gibbs, of Auckland New Zealand. Mr. Gibbs path to system development is a familiar story. He read his first technical analysis book when he was 20, and has been hooked ever since, devouring as much knowledge on the subject as possible. A user of TradeStation since it was known as Omega Supercharts, Andrew has become self taught in system programming.

Andrew attributes much of his knowledge to Larry Williams, whose book, "Long Term Secrets to Short Term Trading" has some fantastic ideas according to Andrew. He is quick to point out, however, that the book could be the best book ever written on data mining (curve fitting), so care needs to be taken.

Mr. Gibbs has a bachelor degree in Business Economics and Accounting, and a Post Graduate Diploma in Applied Finance and Investment. He currently works for Goldman Sachs in one of their New Zealand offices.

How does it Work?

AG Xtreme follows a volatility breakout logic, but is unique in that it utilizes a volatility cycle and chart patterns on a dual time scale to trade only in periods of low volatility where the odds of a large range day are increased.

The idea of using a smaller time frame as opposed to traditional systems that use daily bars came from extensive research that an earlier entry could be attained and drawdown could be improved by breaking the day into two parts, a morning session and an afternoon session.

The other consideration for breaking the day into two parts came from the fact that volatility breakout systems are generally not able to buy below the open of the day or sell short above the open. By separating the day into two parts, if a rally occurs in the afternoon or perhaps a selloff, the system logic is able to either cover the existing position should the retracement be large enough or even perhaps reverse the position to take advantage of a possible rally in the opposite direction.

The parameters for the system to stop and reverse are rather strict, however as the systems main goal is to capture the larger range days that open near the low(high) and close near the the high(low). It is the potential profit of those days that pay for all of the choppy days that finish with the system either making a small win or small loss.

Because AG Xtreme looks to profit from large single day moves, traders utilizing AG Xtreme are in effect long volatility. Periods such as those we have seen in recent years (form 1997) where markets have been abnormally volatile are the periods where AG Xtreme is able to take advantage of the many low risk trades the market throws at us. During these periods the system is far more selective about the trades it makes, and accuracy and trade efficiency is at it's highest during these periods. During periods of lower volatility, such as the past few months, the system actually trades more. Losses are generally smaller, but so are potential profits.

Attain Comments?

There have been periods in the market where volatility breakout systems have performed poorly over a period of 2-3 years. Typically these periods have been during trending markets where daily ranges are narrower as the market "crawls" up or down instead of choppy market movement (on a daily basis) showing large daily ranges and follow through.

AG Xtreme attempts to filter out these periods somewhat with their dual time scale and probability filter, and the testing is very impressive. The hypothetical testing shows the system holding up very well during 2003, when many volatility breakout systems fell on hard times (remember LTS Breakout). It also shows unbelievable triple digit returns in each of the first four years in its back testing. While these do seem too good to be true - if the system can generate returns even half that good you will want it in your portfolio.

Whether the future will bring as good of returns to AG Xtreme remains to be seen. The system is in its fifth month of actual trading at Attain and has performed within expectations, with a very impressive 20%+ winning month followed by a less than spectacular 20%+ losing month. More importantly, however, the returns are lining up with the hypothetical testing in real time. This validates the backtesting somewhat.

It is also interesting to note that the hypothetical returns have decreased every year since 2000, right in line with the decreasing volatility over the same time period. We remain at near record low volatility levels, with many so called experts feeling there is nowhere to go but up. If AG Xtreme is indeed one in the same with being long volatility as the developer suggests, going long volatility at these lows by trading AG Xtreme is an intriguing investment.