First Backtesting results

davidh1819

Member
Messages
74
Likes
2
Hi

Im am still relativley new to trading and I have made a trading system using EMA crossovers on a daily chart and backtested it for 1 year on 6 diffrent markets with the system tweaked to suit each individual market so each diffrent market has diffrent risk reward ratios and position sizes ect. This system was designed with spreadbetting in mind. Its also a long term system with positions lasting a few weeks or even more than a month

here is how the paper trade turned out with an imaginary £2000 spreadbetting account

Backtested one Year plus all the gains and minus the losses

The FTSE 100 made me £720 after 11 trades
The Gold market made me just £80 after 9 trades
The Cocoa market made me £200 after 7 trades
The Soybean marked made me £80 after 15 trdaes
Nestle plc made me £660 after 13 trades
And the NASDAQ make me £960 after 9 trades

Just a varity of markets i chose

Thats £2700 from £2000 in one year thats 145% return in one year

what are you views on this is my system two risky to be making such large gains?
 
1) It is backtested so you will only know if it is good once you forward test it.
2) Even if these were forward tested results there are not really enough trades to be sure. One year of backtesting on a daily chart is not very much. That's the equivalent of me testing a system on a 5 min chart for one day> not nearly enough.
3) Why did you choose those? Was it because they had good trends in the time period, or because they worked better?
4) What was the maximum drawdown?
 
1) It is backtested so you will only know if it is good once you forward test it.
2) Even if these were forward tested results there are not really enough trades to be sure. One year of backtesting on a daily chart is not very much. That's the equivalent of me testing a system on a 5 min chart for one day> not nearly enough.
3) Why did you choose those? Was it because they had good trends in the time period, or because they worked better?
4) What was the maximum drawdown?

2) Yeah good point i might continue the backtest back a few more years
3) they are mainly random there where others i chose but couldent affort to trade them because the pounds per point where to much to risk
4) im not shure i havent added any dates into my spreadsheet but i am risking 2-4% percent of my account on each trade

just give you a better idea here is the FTSE 100
risk reward ratio is 1/3
stop loss 80 points profit capture 240 points at £1 per point
open position when 10 day ema crosses the 20 day ema

£240
-£80
£240
-£80
-£80
£240
-£80
-£80
-£80
£240
£240

total £720
 
You definitely need to backtest longer, over periods that include different market types, e.g. ranging, trending.
 
You definitely need to backtest longer, over periods that include different market types, e.g. ranging, trending.

Absolutely. Trending markets are great for systems which employ MA crossovers. They tend to bleed money horrendously when markets are rangebound.
 
I agree with both Adamus and Jbat. You should also consider checking robustness of your system by varying its parameters in a narrow range in your backtests. The first step, though, is certainly to test against various market conditions as they suggested.
 
I agree with Jbat....the momentum model I use keeps me out of chop...as it does not perform well during chop/range bound. Unless.....I use tick charts and change my time frame....then It becomes a matter of being quick enought to be able to take the setups and signals (I trade the S and P emini and when it is choppy it usually presents a four point range after 20 minutes or so....I am able to make 2 sometimes 3 trades in this range until it creeps out of it....Have a look at the emini historical data on a 512 tick chart and when a momentum move ends....you will see the 4 point give or take a few ticks range.

Cheers
 
Top