how much period of data i should backtest a strategy for intraday in 5 mins timeframe? and i want to do backtest manually.
1 Like
I Personally do back test for 6 months to 1 year. some strategies will not be favourable when you go beyond 1 year+. Atleast 6 months is mandate for a good strategy, If a strategy gets max number of profitable trades in last 6 months thats enough.
2 Likes
How do you back test? With code or Streak or any other website?
2 Likes
at least 7 years.
you cannot manually backtest. use python or c++.
The reason for this one or years will perform well. but over a period of 5 yrs or more and you come to know the issues in your strategies.
The results will be shocking to you.
Also add in the cost of trading + all expenses.
1 Like
Any websites for those who cannot use programming languages, apart from Streak?
manual, am not good at coding. I don’t find streak useful to me
they say market move in cycles, i call it a season, so the season changes after a specific period of time.
you wont get anything useful or reliable If you test any strategy for years. none will be reliable. this is my opinion.
like they say follow the trend, here you have to follow the season, every season has a specific character. so following the nearest would give you good results.
3 Likes
Can you explain, how do you that? I would like to know how it is done.
1 Like
for ex, if my strategy is observe the 1st 15 min candle and I’ll buy if market goes above the 15 min and sell below the 15 min candle means. i’ll change the time frame to 15 min. and then will see each and every day charts for the last 6 months to 1 year. 1st I’ll calculate probability of profit, if it is successful 7/10 time i confirm that is a successful strategy, below that i wont consider, since its momentum strategy i dont consider risk reward here, only pop.
This will take upto 2 hrs to do manual, but I love reading charts so I do this everyday i dont get tired.
above said is an example, not my strategy or advice.
3 Likes
Sounds time taking and hard, as it is manual.
So in the example, you will check the 1st 15 minute candle for the past 6 months, and see how many times the next candle has gone up or down. So, if there were 120 trading in the past 6 months, you check if the price has moved up the 1st 15 minute candle, or down the 1st minute candle, and if the price has moved 84 times, that is 70% of those 120 days, you will consider this as a successful strategy.
So you make all the 120 entries, either up or down, manually.
you don’t need to make any enteries, just looking at the chart and use of some trend lines you can backtest. for some one new it might take time, for me it will hardly take 30 mins for 6 months backtesting.
and entry will definitely not be in the 2nd candle, entry in 2nd candle mostly will fail, you have to enter in retracement in this type of strategy,
above is an example, not my strategy.
2 Likes
Forget about the strategy, but how do you back test is my question? Because you don’t code or use websites. You have a strategy that you want to test, a 15 minute candle or a day candle, what do you do?
Don’t mind my asking like this, I have not heard about people back testing manually, so I am asking.
youtube. tons of tutorials on coding and trading.
streak is designed to get dumb retail to day trade more to generate brokerage. It’s not designed to make you profitable.
I don’t know anything about coding, cannot learn now specifically for trading, I would rather learn more about charts. The discussion started as it was manual back testing.
But Streak has automated back testing. If you are a trader like your ID suggests, what do you do, code?
1 Like
No bro listen
For manual backtesting go to site:gocharting . Com
Lets say you want to backtest from January to till today
Select your time frame (15min)
Now go to January by swaping the chart left to right
Now just select bar replay option above
Lets say bar replay turned on from January 5th then from jan 6th to todays date all candles/data will be erased
Now you can play/pause the bar replay option and check candles one by one
And if you find any trade just note it down on excel sheet or hard copy it…
Test for atleast 6 months of data as said above and you are good to go…
1 Like
Tnx bro all confusions cleared…
2 Likes
Tnx bro
1 Like
u need to whatever is required to become a trader. if that requires you to learn coding, then you learn coding.
u cannot backtest manually for 7 years. six months is not sufficient to tell you if your strategy is good or not. even 1 year of backtesting is not good enough.
if u cannot learn coding or its not your cup of tea,then you can hire coders to do the job. There are a ton of freelance coders who can do it for you.
streak is not good enough.
Also , you do not have to reply to me. Feel free to ignore me.
3 Likes
Everyone have their own way of testing trading etc. You won’t get the comfort in automation or algo like in manual. I always prefer manual. anyway the questionnaire have mentioned a website right, you can follow him.
2 Likes
Ideally you should be backtesting with data that covers different market conditions - up, down, flat, volatile to get a hang of how your strategy is.
1 Like
Not that coding cannot be my cup of tea, it could be, but it does not serve any other purpose for me, its application for me is limited.
I am learning, so I will not ignore.