Next Generation of Trading Systems- Looking ahead to 2006:
December 27, 2005
Come the end of any calendar year it is human nature to reflect on the past and look ahead to the future. This week I have the distinct pleasure of touching briefly on the past and more importantly introducing you to the next generation of trading systems and trading system portfolios.
Trading systems in their raw form have been around since the beginning of the markets and have been fortunate to evolve with both time and technology. It wasn't until the late 80's to early 90's when, for the first time, investors began to hear of quantifiable methods for back testing and later investing in the markets on a regular basis with a strict set of rules. The key to this equation was the computer and its ability to quickly test the validity of an investment strategy over many years of historical market data. At the time most strategists or system developers only had access to end of day market data whereby to basis their strategies success; however as we are all aware of today is a much different story. Today CNBC broadcasts time in hundredths of a second along with market news close to 24 hrs a day, global exchanges have extended their trading hours to meet liquidity demands, and more than 50% of global exchange trading volume is done electronically. All this is to say that market data is now more readily available then ever which in turn has led to the ability for system developers to research and build trading system on much shorter time frames and in many more markets.
As a result of improved data Attain Capital is able to offer investors a range of investments covering a wide variety of time frames and markets. Of the markets available 90% or more are traded on US exchanges— this is about to change.
Over the course of the past 10+ months Attain Capital has been laying the ground work to expand our trading operation overseas whereby US investors will now have direct access to a full range of end of day and intraday trading systems in the most actively traded European futures markets — including the German Dax, Euro Bund, Euro Stox50, and more .
Through our strategic execution partnership with TradingMotion (www.tradingmotion.com) our investors now have exclusive access to a range of systems designed specifically for the European markets via their existing Attain account...New investors are welcome.
Many of you might be saying, or at least I hope you are saying, hold on just one second…tell me more about these systems and how they work. You are in luck- below is an introduction to the next generation of your portfolio.
Before we learn about the specific systems it is first important to understand more about TradingMotion. TradingMotion is the #1 futures execution firm in Spain with over 4000 clients and in an effort to offer their clients a full range of investment products began offering European-based trading systems around the turn of the century. Since then European markets have exploded with volume and their system business has flourished.
Similar to Attain, TradingMotion firmly believes in 100% mechanically traded strategies across many different time frames and system logics. As such TradingMotion works with many different 3rd party developers whose expertise and focus is on the European markets. While there are likely hundreds of European-based systems TradingMotion has maintained the goal of working with only the best.
AutoTradingBot is one such developer and will be the focus of the remainder of the article.
AutoTradingBot Systems:
The fundamental characteristic of AutoTradingBot's systems is simplicity. Through their research they found that using fewer parameters per system resulted in a reduced chance of over optimization. Further, all of their systems have market driven entries coupled with adaptable exit strategies designed to protect and lock in profits when markets are trending in the systems favor.
"To incorporate too many parameters contributes to the potential for our system to experience an excessive adjustment to the historical curve of prices, (i.e "learning" too much of the historical data)."
As with any system, the key to the long term success of an investment strategy is consistency coupled with out of sample performance- AutoTradingBot's systems have both. Noting that when a trading system is first developed it has a tendency to be optimized to the sets of available data during the time of development AutoTradingBot implements a method whereby they offer clients the opportunity to invest in systems that have either fixed parameters or time adjusted parameters. A fixed parameter system is one that uses the exact same logic today as it did when it was first developed; while time adjusted parameter is one that is optimized over time.
Clearly any type of optimization has its risks - will the new parameters reduce the systems average winning trades, will it increase drawdowns, how do I know what type of drawdown to expect with the new system, just to name a few. We have all seen the pitfalls of optimization; however I will argue that most of the negative connotations associated with revising a system come from the fact that revisions are typically made only when a system is in drawdown or the developer "feels" like a new input will work better in the existing market conditions (i.e a new low volatility filter or moving from a 1 minute chart to a 5 minute chart). AutoTradingBot's systems are very different in this sense because they actually use time as the determining factor for when to re-optimize a strategy.
Optimization of their strategies occurs at different times and is conveyed to investors in the strategy name. For example below in the table below "Epsilon 12/12 Bund" uses 12 months of data ending December 30th to determine the trading parameters for the next 12 months and then repeats the same exercise in each of the following years for the year ahead. Another type of optimization utilized is historical optimization, in which they fix the beginning of the optimization period and continue to add new data. For example "Rho H/3 Dax" begins tracking data in January 2001 and then re-optimizes the entire data set every 3 months.
Optimization Examples:

Further differentiating their systems from the traditional development techniques, AutoTradingBot displays any optimized system results on a walk forward basis. That is to say that any optimized parameter is then recorded as out of sample trading and all past results remain in tact with the parameters that were used during their out of sample period (see results below using the table above).
The last and final question is how much capital do I need to get involved? In order to objectively calculate the estimated minimum capital requirements for each system AutoTradingBot utilizes a weighted average of the following 4 parameters:
1. A value equal to 2% of the maximum historical loss of each system. For example, if a system has had a maximum loss of 300 €, the first amount to consider would satisfy the following equation: 300 / 0.02 = €, or 15,000 € per contract. This way, the maximum loss per trade of the system should not exceed more than 2% of the initial investment.
2. The standard deviation of the daily results of the system, multiplied by a factor of 7 (this data point has been referenced in various trade magazines including "Stocks and Commodities").
3. The standard deviation of the monthly results of the system, multiplied by a factor of 4 (this data point has been referenced in various trade magazines including "Stocks and Commodities").
4. The greatest historical maximum drawdown of the system multiplied by a factor of 2.5.
In conclusion, global markets are becoming more and more accessible by the day; however the opportunity for everyday investors to reach them is still very much limited to high load mutual funds or those willing to bun the midnight oil to catch the European market opening @ 2:00 am CST. Through our relationship with TradingMotion our investors now have the ability to expand their investment portfolios across oceans and time zones into some of the most active markets in the world.
To view the performance of the current list of TradingMotion Systems log into www.attainaccess.com and then select "View Computer Backtesting" or e-mail us to receive a performance report @ invest@attaincapital.com.
- Jeff Eizenberg
IMPORTANT RISK DISCLOSURE
Futures based investments are often complex and can
carry the risk of substantial losses. They are intended for sophisticated investors and are not suitable for
everyone. The ability to withstand losses and to adhere to a particular trading program in spite of trading
losses are material points which can adversely affect investor returns.
Feature | Week In Review | Chart of the Week |
.gif)
Feature | Week In Review | Chart of the Week |
Around this time last year, we enthusiastically wrote about the "Santa Claus" rally that helped drive equities to new highs. Unfortunately, the anticipated rally did not fully materialize last week as many investors appeared to be taking money off the table instead of putting year end bonuses into stocks. For the week, Emini Midcap futures gained nearly 1 %, Emini Russell + .8 %, S&Ps + .2%, Nasdaq - .55% on extremely light volume.
Although equities were sluggish for the week, energy markets picked up the slack and gave traders more than enough opportunity to profit from the move to the downside. Natural Gas was the biggest mover falling 9.6 % on the week. Other energy markets traded lower as well but much more modestly with Heating Oil down 2.1 %, Unleaded Gas - 1.4 % and Crude Oil -1.05 % all in the February delivery.
Grain markets were uncharacteristically volatile last week with Soybeans up 3.73 %, Kansas City Wheat + 3.46 % and Corn + 3.13 %. This may be the start of a trend reversal from the huge declines seen in the grains throughout the year.
***Day Trading***
Lackluster trading conditions kept many day trading systems on the sidelines for the week. Those that did participate found it tough to overcome the choppiness and ultimately finished in the red.
Two systems that were able to escape the week profitable were Clipper eRL and Electric Daybreaker II ES. Clipper eRL had three trades amounting to gains of +$180 per contract. Electric Daybreaker II ES traded twice for profits of +$40 per contract.
Several systems merely tested the waters and escaped the week with small losses. Electric Daybreaker NQ had just one trade for a loss of -$10. Impetus eRL also had just one trade from Monday that lost -$64.50. Spectrum SP and eRL both went short on Tuesday and lost -$75 and -$130 respectively. Daybreaker SP lost -$325 on a long trade from Wednesday. Finally, R-Mesa eRL lost -$370 per contract.
Helix ES had a tough week with losses of -$692.50 on half a dozen trades. RC Success ES has yet to return to its winning ways and lost -$785. R-Mesa SP had two trades on Wednesday for a total loss of -$1,284.48. Compass SP also had two trades for a loss of -$1,325 for the week though it fired back with a small winner on Friday to end the week on a positive note.
***Swing Trading***
Several sharp trend reversals sent many swing systems spinning their wheels-especially those that work off intraday bars. In general, the systems that limited their trading performed better than those that were actively reversing their positions.
Bond systems were red hot last week with Jaws Narrowneck portfolio making +$3,150 including +$2,025 in closed out profits. Two of three time frames entered long last week on Thursday and caught the majority of the move to the upside. Mesa Bonds added +$1,156.25 in open trade profits to its current long position as well.
Tzar limited its trading and by doing so outperformed most of the swing systems that trade the stock index futures. Tzar reversed long in the eMD, eRL and ES for profits of +$1,640, +$1,465 and +$527.50 and reversed short in the NQ for a loss of -$616 on the closed out trade. SeasonalST ES also went long last week and was up +$325 per contract in open trade profits.
The Axiom Index programs struggled with the choppy trading conditions last week. Axiom ES lost -$1,277.50, eMD -$1,160.40, eRL -$560 and the NQ +$60. The systems found a nice bullish trend towards the end of the weak to add some open trade profits to the new long positions.
Delphi had similar results to the Axiom programs. Delphi eMD lost -$1,049.50 while the eRL lost -$170 on closed out trades. Both systems were long heading into the weekend. Eclipse eRL suffered the largest losses of the swing systems down -$3,340 on closed out trades.
***Long Term***
It is no secret that several trends that have been profitable for long term programs this year have started to unwind as we head into the New Year.
Some recent examples have been in markets like Corn and Soybeans, which were very profitable for trend followers in '05. Last week's continuation of the developing bullish trend in grain markets had long term programs exiting short trades and initiating new long positions. Andromeda exited its short Corn position for a gain of +$1,337.50 on two legs of the trade. Brix entered long in the Kansas City Wheat and was making +$400 in open trade equity by the end of the week. Brix is also holding long in the Soybeans from a trade it opened last week.
Some other trends showed signs of slowing but regained some momentum late last week. Foreign currencies were up against the Dollar early in the week stopping some systems out of short foreign currency/long Dollar positions but regained their downward momentum later in week. Aberration Plus exited its long Dollar Index position for a loss of -$1,295 on the closed out trade only to watch it rally over 1 % later in the week. Axiom LT also entered short in the Swiss Franc as the market broke lower.
Other markets have been making new highs/lows on a weekly basis and don't appear to be reversing any time soon. Sugar added another 3.02 % last week hitting new all-time highs. Systems holding long Sugar positions include Aberration Plus, SEMA4 Symmetry and Axiom LT (London Sugar).
Please Login to: http://www.attainaccess.com for the latest updated statistics.
IMPORTANT RISK DISCLOSURE
Futures based investments are often complex and can
carry the risk of substantial losses. They are intended for sophisticated investors and are not suitable for
everyone. The ability to withstand losses and to adhere to a particular trading program in spite of trading
losses are material points which can adversely affect investor returns.