The Dangers of Backtesting Forex

1. Start trading tomorrow with real money. –It’s the fastest way to know This is a very painful strategy, but who knows, you just might be a genius.
2. Back test your system with historical data. – That’s ok if you believe life is repetitive in nature. Many a professional trader swears by this. I don’t but that’s just me.
3. Forward test your system. – This is boring and hard work. However, I know it works all the time.

Whatever you do is not the point of this post.

I would like to discuss the problems you should watch out for when you are back testing.

1. Not Back testing- Ok I have said I don’t believe in back testing, however if you don’t back test, how sure are you that your trading system has ever made money before? By not back testing, it’s like you going to buy a house in a neighborhood that has thugs, just because you like it. How will you know that your system can make money?
2. Not knowing how to evaluate your results. – Every time I go to my dentist, they show me some x-ray things that are meant to mean something to me. I always want to ask the doctor how he expects me to know there’s a hole in my tooth when all I can see are black and white objects in the x-ray. You had better know number of trades, profit per trade, consecutive losers/wins, and most important, your drawdown. I don’t care what system you use, there will be periods you are going to lose money. The question is, can you afford the loses? Now compare your systems and find one that YOU can live with.
3. Be a bit negative. – I am a very excited guy and every gal I loved I had to bring to my mom. The reason was simple. My mom would question her better than I could. You had better find a way to question your system. If it means posting it on forex forums so that the trolls can disappoint you, you do that. There is nothing worse than over confidence and the forex market has no mercy for over confident fellas.
4. Ignoring Commission and slippage. – It is not true that there are no commissions if forex. Every time you place a trade you broker is getting something and it’s a cost for you. A simple example is comparing oanda to fxcm. Oanda has 2 pips on the dollar and fxcm has 5 pips. That’s a difference of 3 pips. If I trade one lot every day for a year, that means I am losing $ 20 per trade times 20 days a month which is $400 per month and $ 4800 every year without even thinking of my trading system. Now if you don’t think $4800 a year is important, you are a rich man and should not be trading. Your trading system should take note of slippage and commissions.
5. Using bad historical data. – By the time you got to this site, you have gone to all those sites that like back testing a system. Funny thing is that I’ve noticed that once the system is online with real money, it quickly disappears. This especially happens with metatrader based systems. The reason is very simple. As much as your broker will tell you that demo data is the same as real data, they are lying. Next metatrader data is not the same as fxdd, or northfinance data. So while you were busy testing your next Holy Grail forex system, the historical data from one broker to the next is different. I have tried for so long to explain this to my students but it never gets anywhere. You cannot back test a system with fxcm data and then trade it with northfinance real accounts. Remember there’s no single clearing house in forex so whatever data you are getting is cooked. Try and find a neutral source for you data and back test your system. Then get proper sample.