BACKTESTING - How to Start?

[robin73736 10]

Hi, is there a very easy way to test simple strategies?

I just want the easiest possible way to play around with some ideas.

 

Thank you,

Robin

[Tradelikespock 10]

Need more info like time frame you're trading. Ninjatrader records market price movements for you and it's great for day trading (which I magically use for trading weekly charts lol). You can also purchase historical price movements on CD but I forgot where you get those.

 

You could get sim accounts too but IMHO trading sims and backtesting will never be as good as real world testing.

[prax 10]

you have to spesific your question by adding manual trade backtesting.

many people would think, it's an EA backesting. for ea backtest you can start with strategy teste on metatrader. for backtest manu trading plan, i use a script on my Armada Markets platform called "LFH trading simulator" it's a simulator for your manual trading strategies.

check attachment for the file.

armada markets_Trading_Simulator.zip

[shuvo7748 10]

I also want to know ore about that. anyone please tell me more and where i get more info about that.... :missy:

[pro4Xtrader 10]

At a time I did a lot of backtesting on various platforms.

 

The perfect equity growth turned into free-fall almost immediately and almost always.

 

Best results can be achieved by backtesting and getting out of sample date on most recent data. Frequent re-backtesting every x-/days/weeks/months/years (depending on the timeframe you trade) is preferable.

[mangolassi 11]

Automated backtesting any given strategy isn't a good idea unless you are going to be using that same automated system in live trading.

 

In my opinion, if you are manually trading, manual backtesting is the way to go. Since you will be trading with your own discretion, I don't find automated backtesting useful, and even deceitful. You won't trade like a computer in a live environment, so you need to test in a way that most accurately imitates how you will trade or it isn't worth much.

 

When I started trading currencies, I wanted to test some of the price action strategies I was going to be trading. So I just made an excel spreadsheet and put data in it. For example, I took a strategy and started like 2 years in the past on the 8 hour EUR/USD chart. Whenever I'd find it, I'd go through the strategy and find out if I would have won or lost, and how much I would have won or lost. You can plug this data into excel and calculate a whole range of different statistics. You will not only get useful data, you will also learn to become an excellent chart reader with a great feel for how the market you are trading moves. You can't get that sort of incredible utility from automation.

[pro4Xtrader 10]

Automated backtesting any given strategy isn't a good idea unless you are going to be using that same automated system in live trading.

 

It appears that you are saying: "ok I am going to backtest this strategy, but I'll use another one for live trading". What is the point?

 

In my opinion, if you are manually trading, manual backtesting is the way to go. Since you will be trading with your own discretion,

 

Automation is always preferable since it just makes all calculation easier. If by "discretion" you are referring to partial trading without using rules, then you are gambling.

[mangolassi 11]

It appears that you are saying: "ok I am going to backtest this strategy, but I'll use another one for live trading". What is the point?

 

I didn't say that - I basically said that if you are going to backtest, it would make more sense to test in a manner that you would be trading. If you are a discretionary trader who is not using automation to execute trades, you should backtest your strategy by chart reading manually. That makes most sense to me, and I think a lot of people would agree with that.

 

 

Automation is always preferable since it just makes all calculation easier. If by "discretion" you are referring to partial trading without using rules, then you are gambling.

 

 

By discretionary, I don't mean that you don't have a plan - it means that you have the ability to use some subjective reasoning to read market context and setups. The best price action traders, for example, have worked out their setups in a pretty black and white manner, but knowing when to execute the setups and money management can be a highly subjective thing based on experience and individual skill.

 

It's extremely difficult to automate strategies - it is not a simple issue of making a computer do A when B happens. Trading is far more complex than that. For that reason, it is going to be very tough to be able to rely on automated backtesting, unless you ARE going to be using automation and are going to be trading with that same automated strategy that you just tested.

[zdo 20]

Hi, is there a very easy way to test simple strategies?

I just want the easiest possible way to play around with some ideas.

 

Thank you,

Robin

 

Have you looked at TradeStation?

It's not necessarily "easy" (for everyone) - but it still may be the "easiest" way to prototype simple strategies

 

hth

[divyanshisharma 2]

Best way to know this information about backtesting!