For new traders, backtesting the day trading system is a must. Markus, an expert in day trading, shows you how to backtest correctly.
Backtesting can be a very powerful tool if it is used the right way, in my opinion.
The problem is that many traders use all of the backtesting software's features too much because they think that more is better. Many so-called system developers try to make it sound like the longer you backtest, the better and more stable your system will be. Not always, though.
Let me show you the e-mini S&P. In 2000, the average number of ticks per day was between 100 and 150, but in 2004 there were only 40 to 60 ticks per day. If you backtest any trend-following day-trading system (http://www.rockwelltrading.com), you can see how well it did in the past. In the e-mini S&P, you can see that it worked perfectly until 2002, when it all of a sudden broke down. It looks like there are no more trends during the day. That's not surprising, since the e-mini S&P's daily range dropped by more than 50%.
What went wrong?
There are a few reasons why. The Pattern Day Trading Rule, which was put in place by the NYSE and NASD in August and September 2001, is likely the most important: If a trader makes four or more day trades in a five-business-day period, he must always keep at least $25,000 in his margin account. Because of this rule, traders stopped daytrading stocks online and instead started trading the e-mini S&P future.
Look at the sudden rise in the number of e-mini S&P trades at the start of 2001:
Many of these day traders in stocks used strategies to make a few cents by "scalping" the market. Using the e-mini S&P, they suddenly had a lot more leverage, paid less in commissions, and made a lot of money.
Unfortunately, these scalping methods kill an intraday trend almost instantly, so almost every trend-following strategy fails.
The launch of automated online daytrading strategy execution in TradeStation was another big reason why the market changed so much. In 2002, there were 268 percent more TradeStation customers who used this feature. Overbought/Oversold strategies became very popular, and when the market tried to go in one direction, these online daytrading strategies set up a position in the opposite direction right away.
Conclusion
You need to know these things about backtesting. It's not enough to run a system on as much data as possible; you also need to know what's going on in the market.
You should use trend-fading systems in markets that don't have trends, like the e-mini S&P, and trend-following systems in markets that do have trends, like commodities. And that's when backtesting comes in handy:
If your backtesting shows that a trend-following method worked from 2000 to 2002 but doesn't work in 2003 and 2004, you shouldn't use this strategy right now.
And vice versa: if you see that a method for going against the trend made you good money in 2003, 2004, and 2005, trade it.
day trading online
The strategy that works no matter what is going on in the market, whether there are trends or not. Most of the time, a strategy works very well in ONE market condition (like when it's trending) and causes small losses in the other. Because of this, daytrading strategies need to be changed. Backtesting can help you.
