How to analyze backtested data?

A

anituchka

Hello, I have developed an automated trading strategy that has shown good results in backtesting. I am trying to
do some additional analysis and possibly try to find any correlation but am a bit stuck for ideas.

I have the following data.

Average true range (in ticks)
Opening range size (in ticks)
Daily profit (in ticks)

Does anyone know how I can analyze the data? I have about 850 rows...Ideally I wanted to know if there is any correlation between average true range and opening range or opening range and profit or profit and average true range.

Any input will be appreciated.
 
What type of instrument: stocks, commodities, forex, ...
Opening Range Size : over how many ticks ? over how many minutes?
Do you have an histogram of it?
 
I assume you are using excel.

Can you put in a chart of :
(prior day ATR, next day Opening Range on 30minutes).
Keeping a plot with dots, so as to see where/IF there is a high density of dots.

( actual day Opening Range on 30minutes, actual day ATR).
Keeping a plot made of dots. Same reason - IF important.


Your Daily profits : are these your trading profits ?
Your trading based on opening range breakout?
 
Search for a youtube FuturesTrader71 Statistical Analysis for a Broader Edge. About 18 minutes in he shows how to do a correlation, statistical analysis, live in Excel with imported ES data.

Simplest approach because you get to see step by step.
 
Quote from anituchka:

Hello, I have developed an automated trading strategy that has shown good results in backtesting. I am trying to
do some additional analysis and possibly try to find any correlation but am a bit stuck for ideas.

I have the following data.

Average true range (in ticks)
Opening range size (in ticks)
Daily profit (in ticks)

Does anyone know how I can analyze the data? I have about 850 rows...Ideally I wanted to know if there is any correlation between average true range and opening range or opening range and profit or profit and average true range.

Any input will be appreciated.
More...

For quick visual inspection:

- sort your data for 1 criteria (eg, Opening range size, lowest to highest)
- add a column which will be the running sum of the Daily profit

Graph the running sum of the Daily profit ... the "good" values for the criteria correspond to the part of that curve which are going up (and possibly, think about doing the opposite on the parts of the curve going down).

Attached is a Time-of-Day analysis for one of my setups.
 

Attachments

Quote from anituchka:

Hello, I have developed an automated trading strategy that has shown good results in backtesting. I am trying to
do some additional analysis and possibly try to find any correlation but am a bit stuck for ideas.

I have the following data.

Average true range (in ticks)
Opening range size (in ticks)
Daily profit (in ticks)

Does anyone know how I can analyze the data? I have about 850 rows...Ideally I wanted to know if there is any correlation between average true range and opening range or opening range and profit or profit and average true range.

Any input will be appreciated.
More...

in python ... you would need to change input ... pd.read_csv('youfile.csv')
Code: 
import pandas as pd
import Quandl
es=Quandl.get('OFDP/FUTURE_ES1')
es.corr()
 
Quote from smallStops:

I assume you are using excel.

Can you put in a chart of :
(prior day ATR, next day Opening Range on 30minutes).
Keeping a plot with dots, so as to see where/IF there is a high density of dots.

( actual day Opening Range on 30minutes, actual day ATR).
Keeping a plot made of dots. Same reason - IF important.


Your Daily profits : are these your trading profits ?
Your trading based on opening range breakout?
More...

Yes, trading profits. Yes, my trading is mainly based on the opening range breakout.
 
You must log in or register to reply here.
Back
Top