SUMMARY

Psychology

• System trading is perfectly designed for proactive trading. You can establish specific trading objectives and criteria, pay close attention to the marketplace, and measure your performance objectively. But in the final analysis, you have to call the trade, and decide when to place your order. You have to take responsibility for missing a trade, or the slippage in executing a trade.

• Systems do not entirely insulate you from your own emotional reactivity to the success and failure of your trades. Realize that all trading is influenced by lifelong beliefs and habits that are operating for you when you trade. To the extent that your trading is dominated by long-standing life principles, you may not be maximizing your opportunities. Instead, you may be limited or stopped by your underlying assumptions of the world.

• You must learn to separate trading events from your emotional reactions so that you can see any habitual behavior that leads to compulsive and inconsistent trading.

• Record your trades and after the session has ended, review the facts and how you responded. This way you can see how emotions may have clouded your performance, leaving your mind clear for the next trading day.

• Enter into a meditative state of consciousness to reduce anxiety and tension and decrease the impact of distracting stimuli. It helps you distance yourself from the sequence of emotional responses and negative interpretations of stressful and unsuccessful trades that trigger panic, denial, rationalization, and avoidance-all of which perpetuate the circle of failure.

• Make a conscious commitment to a financial objective greater than you are accustomed to achieving. This means relinquishing any inclination to trade in terms of comfort and safety, and to allow your choice of price, volatility, and size of positions to be guided by parameters commensurate with your goals. This will help you overcome a natural fear of change.

• Do you have what it takes to succeed? Ask yourself these questions:

-You will need to develop market intuition, which only comes from trading experience and a fair amount of reading. Your initial investment combined with initial losses while trading may easily exceed $5,000. You may need to wait a year or so before you break even. Do you have the financial resources and are you patient enough to prevail?

-Proper system development will take longer than you want it to. Develop-test-develop-test. . . Can you afford the time to do a proper job and are you patient enough to avoid premature risk taking?

-You will need to be fluent in handling hardware and software bugs and breakdowns (e.g., database corruption, incompatible software, hard drive failure).

Can you afford the time to learn all this and are you patient enough to learn frustrating tasks?

Risk-Exposure

• Select exposure duration thats best for you. Although position trading (holding your position overnight) typically yields more profit than day trading (exiting your position at the end of each day), it is more important for you to select the method you can live with in terms of stress, risk, required skills, and required capital, than to make a choice based on expected annual profit.

• You can control risk by working in a time frame whose average risk, as measured by the KAR or VAR method, is just under the maximum risk you are willing to accept. For example, consider wheat futures, where 1 cent is equivalent to $50 risk exposure. A 7 cents average risk translates to $350 of exposure per contract.

• By shortening the charts time frame, risk through price action drops but commission costs per trade increase in importance. One way to handle this balancing act is to initiate trades using a short time frame, a 15-minute bar chart for example, and then lengthen the time frame while in the market.

• When day trading, tick volume bars are superior to time bars in most cases because they form more quickly during periods of high activity and more slowly in periods of quiet activity. This produces visually cleaner signals.

• A more obscure aspect to risk control is the notion of "time in the market." It is better to be exposed to the market briefly than for extended periods of time. With all other factors the same, I prefer a system that has me holding positions (exposed to the market) for the least amount of time, because while I am out of one market, I can invest the money elsewhere.

Risk-Diversification

• Portfolio theory extols the virtues of lowering risk by holding well-diversified portfolios of securities or by trading in unrelated futures markets. Although futures are more leveraged than equities, diversification reduces risk faster with futures because unrelated futures markets are less cross-correlated than unrelated equities.

• Diversification can also be attained through multiple systems, each running in a different time frame, or a different set of indicators.

• To construct your optimal portfolio, first remove all contenders that deliver poor reward/risk performance as measured by the Sharpe ratio or by the gain/drawdown ratio. Create a return matrix and covariance matrix and feed them into a mean variance optimization program.*

* There are several low-cost, stand-alone software packages available to do this. For example, Ralph Vince offers Portfolio Manager. More general-purpose mathematics products like SAS and MATLAB can also perform this task.

Part III

Risk-Miscellaneous

• Risk is directly proportional to volatility. You can control risk by placing stops above or below the mean price action by an amount that is scaled to the standard deviation of either the bar-to-bar true range or the bar-to-bar change in closing prices.

• Get comfortable with all the mechanics of trading, including finding the right broker, data vendor, and so on. Then as your performance improves (hopefully), you may want to increase your risk with larger lots, leveraged trades, and volatile markets.

• A winning streak or losing streak can readily change your perception of your systems capabilities, possibly leading to over- or underconfidence. Either way, your system has not changed, you have. To help you stay neutral, analyze your systems winning and losing streak tendencies. If your system is showing consistent losses unlike you have seen while testing, then stop trading and examine the nature of your systems losing streak.

Charting

• "Zone analysis" quantifies how a market has behaved while an indicator was in various zones. Typical zone labels may include "bullish," "bearish," or "neutral." This approach may require additional analysis based on whether the indicator was rising or falling within a zone.

• "Subsequent performance" analysis gauges market performance over specific time periods following a specified condition. This approach is applicable to indicators with signals that are most effective for a limited time, after which they lose their relevance.

• Consider using primary indicators to time your trades and secondary indicators to screen or filter your zones of trading opportunities.

System Development

• Once you develop a satisfactory core trading system, you can add factors one at a time in a stepwise manner. During this process, each additional input should be in tune with what in known about each market. Parameters being optimized should be stable and robust; and, finally, each additional input must hold up under out-of-sample validation, as well as in-sample optimization.

• If you believe your system is making poor decisions, exit the market, stop trading, and rework the system until it once again meets your expectations. On the other hand, avoid being stuck for years developing a system at the expense of real trading experience. You need both.

• Your historical data should be clean and contain as much history as possible. When it comes to data, more is better. If its inaccurate or incomplete, it can do