Creating a trading system from scratch

How many lines of code you are comfortable with


  • Total voters
    61

VJAY

Well-Known Member
Hi
Just created an application to backtest from Excel.
You can access it here
It would be free for the next quarter for all members here.
Presently, you could backtest only for the last year (2018)
Dear UB,
Happy new year!!!what this file?backtest which one?
 

UberMachine

Well-Known Member
General principles. The backtest is based on the assumption that you start with price/volume data, add columns to the data, filter upon conditions and then run them on a universe based on a metric already provided as a column. All the parameters specified are parameters to the backtest functions in fastbt

The worksheet is divided into 3 sheets.
  1. Parameters - general parameters for the backtest including capital, commission, slippage, order type, price to be orderd, number of orders etc.,
  2. Columns - the columns to be added. This must be explicitly specified and it is a good practice since it makes us think in a structured manner instead of trying things arbitarily
  3. Condtions - list of conditions to filter data. They are just strings based on columns already existing in the dataframe
 

UberMachine

Well-Known Member
Think of something similar to a screener but with backtesting (and real time monitoring to be enabled).
This is very much in the beginning stage and its more structured than the available screeners since it forces you to think and structure the way you want to approach a strategy. Would improve upon the documentation in the coming days.
Regarding indicators, presently its restricted to moving averages and RSI (since they are easier to implement for me :))
 
General principles. The backtest is based on the assumption that you start with price/volume data, add columns to the data, filter upon conditions and then run them on a universe based on a metric already provided as a column. All the parameters specified are parameters to the backtest functions in fastbt

The worksheet is divided into 3 sheets.
  1. Parameters - general parameters for the backtest including capital, commission, slippage, order type, price to be orderd, number of orders etc.,
  2. Columns - the columns to be added. This must be explicitly specified and it is a good practice since it makes us think in a structured manner instead of trying things arbitarily
  3. Condtions - list of conditions to filter data. They are just strings based on columns already existing in the dataframe
really great but the backtest hangs and doesn't show results for me
@UberMachine happy new year....any new strategies this year?
 

Attachments

UberMachine

Well-Known Member
really great but the backtest hangs and doesn't show results for me
@UberMachine happy new year....any new strategies this year?
@vinit_7801
Now, you can try it.
There is an error in your formula that might have caused it.
Also, I run the server only 12 hours a day. You can just copy the url so that you could see your results later.
You can try any number of strategies with this tool (most of the general strategies are on the short side but you can try out specific strategies that might give you better results)
 
Last edited:

Similar threads