How can I backtest a strategy?

M

mertkaan

After 2 years of trial and error with different strategies that I gathered from books etc, I have realized (maybe a bit late) that I have to create my own edge/strategy in order to be a profitable trader and also have the discipline to stick to my system during bad times.

However, I can't seem to find a way to backtest my own strategy on stocks & stock options other than manually backtracking through charts which would take forever to do so. The issue is that I have 0 knowledge of coding.

So my question is: How can I backtest a strategy based on TA without any coding knowledge? If this is not at all possible, can someone suggest me a course/education material that would teach me the coding required for a newbie like me to backtest?
 
I've seen and tested some of the platforms for "non-programmers", but I personally hate graphical workflows where seems super easy to code an algorithm by just drag & drop pieces, but after 5 minutes it gets totally messed up. I prefer non coding, non visual editor approach like the one in Arcade Trader. You write sentences in English.

I'm using it for all my alerts and basic algorithms.

For complex stuff I still write code in C, C++, Java, Javascript and Python, but I have to say that the Natural Language approach works (for me). Also, when a client asks me to code an algorithm, if I can use NL it's much easier to explain it to him if he's an old-fashion trader.

Apart from Arcade Trader I've seen other platforms with NL but they are limited to alerts or very basic stuff. Since Arcade Trader is free for private user, I'd go with it and if it doesn't work try something else. If you find anything interesting about other platforms, please share with us.

By they way this is my part time job: coding algorithms for investors/traders. So if it doesn't work for you, I could help by using NL or a programming language.
 
Nakesha said:
I've seen and tested some of the platforms for "non-programmers", but I personally hate graphical workflows where seems super easy to code an algorithm by just drag & drop pieces, but after 5 minutes it gets totally messed up. I prefer non coding, non visual editor approach like the one in Arcade Trader. You write sentences in English.

I'm using it for all my alerts and basic algorithms.

For complex stuff I still write code in C, C++, Java, Javascript and Python, but I have to say that the Natural Language approach works (for me). Also, when a client asks me to code an algorithm, if I can use NL it's much easier to explain it to him if he's an old-fashion trader.

Apart from Arcade Trader I've seen other platforms with NL but they are limited to alerts or very basic stuff. Since Arcade Trader is free for private user, I'd go with it and if it doesn't work try something else. If you find anything interesting about other platforms, please share with us.

By they way this is my part time job: coding algorithms for investors/traders. So if it doesn't work for you, I could help by using NL or a programming language.
More...

Thank you for your input. I will try and get back to you after I'm done developing my strategy. I also found a (paid) service at https://www.getvolatility.com/products/backtesting/ for options backtesting, however I would be glad if someone who has used it before can give some review about it.
 
The stock market is multifaceted. Ones has to have a model that drill deep(detail) and wide(coverage) to predict the market. On top of all these, we have central bank intervention and virus. Once these things go away, the model will unravel again. :(
 
If you wanted to go custom and learn to build your own route you could use python and a TA library. w3schools and stackoverflow can teach you a lot of the basics to get a backtester up and running. Haven't looked in a while but you may find a base to start with at github.
 
Peter8519 said:
The stock market is multifaceted. Ones has to have a model that drill deep(detail) and wide(coverage) ……... On top of all these, we have central bank intervention and virus. Once these things go away, the model will unravel again. :(
More...
%%
200 day moving average helps/bull market+ bear market.
IN 2008 bear market fed cut rates-------- that helped for about 20 or 21 days.; but then bear trend resumed for many months
SO the fed or any can't really ruin it or make me buy.
The virus was a pretty good excuse/hype; until everyone figured out kill RATE was about 1%[11% in Italy/heavy in old age, heavy in smoking Italy =/heavy in socialized med]
Good points.
Mr mertkan; good question +until you are willing to risk a very small amount of money in the market; you will not make money in next 2 years or 20 years.And as far as learning with leverage/options/derivatives= worst way to learn with leverage. SPY or spyg costs aboput $300 or $50, if I could not afford that I would not even try top trade or invest.
Buy/borrow books or a library card or 2.Not a stock/etf tip
 
You must log in or register to reply here.
Back
Top