Mistakes in designing a system can foul up the best traders. Here are some common pitfalls to avoid.

Developing your own trading system can be a rewarding experience. In addition to a growing account balance at your brokerage firm, you will learn a lot about the markets and yourself. During my 15 years in the business I've spoken with thousands of system traders, sharing our biggest successes and our most embarrassing mistakes.

Here are some of the most common mistakes new (and many not-so-new) system developers and traders make.

I. Failing to identify the market condition for which you are developing a system

The most common mistake, particularly among novice traders/developers, is failing to identify what type of market condition they are designing their system for. There are three types of market conditions to consider when designing a system: 1) trending market, where you design a system to catch major moves but it keeps you on the sidelines during choppy markets; 2) choppy, sideways market for which you design a support/resistance system; and 3) markets that explode in short spurts, which profit best with volatilit} expansion systems.

No system I've seen (and I've seen thousands) makes money in all market conditions. The idea is to make money while the market is in one of the three conditions, and not lose money during the other two. Design a system for one of the market conditions and many of your other decisions (such as entry and exit techniques and money management control) will fall into place.

II. Over-reliance on indicators

Don't rely too much on your indicators. Explore entry and exit signals that are designed based on price relationships and price movements.

There is nothing wrong with using technical indicators, but make sure you understand what market condition this indicator is designed to be successful in. First, you should know how the indicator is calculated. If the theory and computations don't make sense to you, chances are you'll misuse it. As a general rule, the complexity of the indicator is inversely proportional to its usefulness and profitability.

III. Too complicated!

Many trader/developers try to use too many price patterns and technical indicators, often in the wrong combination. Keep it simple. I'd recommend two or three market entry techniques, a straight dollar stop loss (which should never get hit except in disaster situations) and one or two market exit techniques.

IV. Curve-fitting

When a new system developer backtests his system and finds any trades that lost money, the first thing he does is go back and examine each loser and write additional trading rules that eliminate it from the results. You should examine the losers. However, only modify the entry technique in a way that makes good trading sense. If you are writing separate rules for every loser, without clearly thinking through the effect this will have on the system, you're curve-fitting.

V. lgnoring the system

A final most common mistake isn't about system developing, it's about trading. Many system developers at some point fail to trade their system. When the system signals a "buy" they override it saying, "No. It doesn't feel right. I don't want to be long today." I call it an emotional override. Invariably your emotional override will pass up your most winning trades. In my experience, my biggest winners were the trades I felt the most uncomfortable about.

Realize that system trading does feel different from system developing and testing. It should. Trading is real money! But remember, you developed your system for a certain type of market and based its trading rules on historical facts, not emotions. When you inject an emotional override, you change the expected results in wildly unpredictable ways, usually for the worse. Remember to follow your system.

Bill Cruz is president of Omega Research in Miami, Fla. Omega's software programs include System Writer, SuperCharts and TradeStation.