Quantitative System Trading
A trading system is a systematic way of trading the market. A trading system is a collection of formulas and rules that generate buy and sell recommendations from price and volume data. The beauty of the system trading is that it does not tie you to the computer all day long.
Systems trading is based on technical analysis of the futures markets, which assumes that a study of the markets themselves can provide an objective quantifiable basis for anticipating future prices. Technical traders create strategies based on their observations of price trends for various futures markets during selected historical periods. Then they convert these strategies into computerized trading systems. Each system utilizes current market information to generate buy and sell signals.
Computerized trading methods rest on the twin assumptions that price trends may be discerned in the futures markets and that speculative risks can be reduced through diversified trading. Even with the use of computerized trend-analysis techniques, however, it is impossible to detect the exact turning point of a long- or short-term price movement. Computers add no "magic" to the problem. They simply enable a very large number of mathematical calculations to be made in a very short time.
The use of computerized systems helps to reduce the level of emotion in establishing and closing-out positions if the system is followed precisely. Also, computerized trading forces the trader to estimate the effects of those components on system performance. Computerized system trading allows one to discover more efficiently and more thoroughly the information locked in the analyst's data. The computer also allows the trader to become swamped with data and to devise systems that explain the past but perform poorly in real time trading.
Many traders prefer using a system because of the rigorous testing involved in their development, and the fact that a system may help to eliminate some of the risks often associated with human error and emotions in making trading decisions. You may find that you prefer systems trading for these reasons. In addition, systems trading also saves you valuable time while enabling you to trade at a high level of empirical mathematical proficiency while having access to institutional service.
Just as there is no best way to trade futures, there is no single "best" system for all markets and all traders. You may want to trade a composite of systems in order to diversify your trading. You can trade a system that specializes in a single futures contract, such as the E-mini S&P 500 stock index, or one that has a balanced portfolio of market sectors, such as bonds, stock indices, agriculture, softs, meats and currencies/FX. There are also professional systems designed specifically for day trading or for longer-term investments. Each system has unique characteristics; using specific money management techniques, operating over a specific time horizon, and having many non-correlated markets in a single portfolio. Also, each system is geared to a specific risk profile.
Trading System Features
Although the actual commodity forecasting process comprises in-depth analysis of all available economic and price-trend considerations, accurate market technical analysis is probably the most important element in successful commodity trading. There are only few precise mathematical definitions of most technical theories. Not only are clear-cut definitions required, but the problem also needs to be defined in such a way that quantitative methods can be used to reach a solution.
A Good Trading System Must:
Be totally objective
Be easy to use
Give clear BUY or SELL signals
Has potential to produce large profits per trade
Attempts to keep
drawdowns to a minimum
Measurement of Return Versus Risk. Commodity trading is a battle between return and risk. Because of the leverage involved, you can achieve a higher rate of return than from most other forms of investment, but at a higher risk. In evaluating a commodity trading system, the relationship between return and risk must be defined. Yearly return, closed-trade drawdown, open- and closed- trade drawdown, profit-per-trade and other metrics each tell you something, but they do not tell the whole story, and sometimes they are misleading.
For example, open- and closed-trade drawdown is the peak-to-trough drawdown in an equity stream composed of the end-of-day account equity figures over some timeframe. This is always the most severe drawdown figure for a system. Some people will look at this drawdown as the minimum account requirement to start trading. This is not true. Much of the “peak” in the peak-to-trough figure is open-trade profits. You would not need the drawdown amount at account origination, because the trader would not have been in those trades.
About Risk. Many traders build their trading plan based on perceived future profits. They look at past profit per trade, or average monthly profits, or average yearly profits when deciding their trading account funding level and the mix of commodities and number of contracts to be traded. This approach neglects the most important decision a trader should address: what level of risk are they willing to undertake? When building a trading plan risk should be addressed first, and the profits that may accrue as a result of that risk level become a by-product. A trader should realistically decide the amount or percentage of account equity he/she is willing to risk before terminating trading and build a plan within those constraints. Traders with a small account should realize that there is a very real probability that they will not succeed and try and determine what the probability of success with that account size is. They can then decide if they are willing to take a 50-, 60-, or whatever-percent chance with those funds. This caution is applicable to our systems and every other trading system/methodology ever developed, because they all have losing trades, and losing periods.
Computerized trading forces the trader to estimate the effects of the components on system performance. A system must be created using a sufficiently long period of time for those markets anticipated as trading vehicles. A trading system that shows positive results for the test period should then be estimated on a "hold-out" sample of data for a period of time after the period used to create the system. If test results remain positive, then trading in real time can begin. The results of any system must include the effect of commissions and other trading costs such as the bid/ask spread. Also, there are a number of aspects of "acceptable" performance that must be considered in judging a system such as dollar return per contract, rate of return on average margin, maximum reduction in account equity and standard deviation of rate of return. While there are different methods for assessing the adequacy of return or the size of risk to which the trader is exposed, the trader should verify, at least, that the return and risk generated over the test and hold-out periods are acceptable and sustainable by those using the system.
The necessity to test profitability over previous time periods leads to a potential pitfall of computerized system trading - optimization. There are a number of parameters of a system - such as the number of days to use in moving averages, the placement of trailing stops, position exit methods, etc. The computational power of the computer enables one to test sets of parameters so as to produce the greatest performance over the historical test period. However, since part of what is being used to identify the optimal parameters is a random component of price change during that period, the same pattern may not repeat in subsequent periods when trading is taking place. Thus, the profits and risk levels indicated by system testing will not be replicated, and a continual round of re-optimization and trading failure can ensue, leading ultimately to the destruction of trading equity.
To reduce this, a system must be chosen based on components that achieve desired effects (e.g. trend-following or trading-range characteristics) or a system with those components that the trader believes for some empirical or economic reason may be helpful in producing profits. The selection of parameters should be based primarily on the objectives of the system (e.g., short moving-average periods for active trading systems and longer moving-average periods for long-term-oriented trading systems). If the components of the system rely on fine-tuning of parameters to produce profits in the testing phase, it is unlikely they will be sufficiently robust to produce trading profits in real time.
If you would like more information about
Systems, please call Toll Free
1-866-424-5826, E-mail email@example.com or send an Electronic Request.