How many trades is enough in backtesting in 5 mins time frame?

Nope, I am trading 2 systems now, both work on backtests of last 10 years. Some years are not so good, some are great, but they work year by year more or less. Works in Forward test too. And combined they do much better than standalone.

I would never trust any result from limited backtest of months or even 1-2 years. That said, if someone can do some sort of walkforward test and keep on adapting to recent changes using discretion, perhaps that could work too, i don’t know. Hard way to work, but trading is hard work too i guess …

If something which is working for years of back testing that s very much good. I dont rely on coding, I don’t rely on automation, I dont rely on algo. I atleast get 1 to an avg of 3 to 5 signal per day from my strategy. It’s just repetitive everyday and 6 months is like 6 years for me. When i said its my opinion it means its my strategy and it works for me.

How is manual back testing done for delivery based short term trades?

When you say your system works, what’s the average returns you get?

Not really interested in disclosing numbers, does not achieve anything.

Its nowhere near twitter type returns but so far have comfortably beaten Nifty in back and forward test every year while maintaining risk/DD.

Main task must be to compound so that say 20-30% today ( that is possible and we can do better ) is less than 20-30% of tomorrow while not risking capital too much, say with 10-15% max DD.
Its not too hard to beat Nifty on returns/DD once you can find edges, but there is always risk that any single system might stop working. There is no guarantee of future performance.

Do you trade stocks or only derivatives? If stocks, is that only day trading or delivery too?

Come on, throw a number. This is not twitter.

This is as real as it gets.

I already did. That’s a decent number for long term returns. I may have done better in last few years but future is unknown. A friend of mine went from about 2-3X that in a year to roughly 0% because his system stopped working after covid. Even 0% needs some sort of an edge due to taxes and slippages.

Rather than fret too much about some imaginary numbers ( imaginary for anyone but me …), best to focus on having some method that allows you to find reasonably durable edges( need some luck here too) + diversify into multiple different systems (makes big difference).

Doing backtesting manually can take a lot of time to learn. You can either use the website mentioned here or you could learn coding, whatever works for you. Go ahead with that. Learning a new skill can take time so you have to be patient.

There are some strategies which can never be backtested through coding.

2 Likes

some strategies, especially mine cannot be automated, hence i always prefer manual

1 Like

Same here. I am a hand trader. And what I know cannot be coded or understood by a machine.

And as far as I have come across, podcast and all, there are lot of traders who don’t need a programming language to be successful.

So, learning a language cannot guarantee your success here. Again it all depends on what you intend to do.

1 Like

Yes as far as I know from many twitter handles and other so called experienced traders who use automation have text book strategies like ORB, Breakout etc and also mainly involves some indicators like moving average, supertrend RSI etc.

These type of strategies when I have both backtested and forward tested have very lest probability of profit and seen many times the stoploss have hit.

having tried almost all so called indicators which lag far behind and was useless in intraday trading i started to trade based on price action only.

Price action trading IMO mostly can’t be automated, even if it can be automated with some parameters it is not completely price action trading.

So for intraday the indicators did not work, did they work for longer time periods, did you check?

I have never traded swing for longer time period since i had a very less capital at the beginning. then moved to intraday. I have used indicators only in intraday. not swing.

I’d advice you to go for price action trading which will help you in swing also. have seen people use moving avg that too for positional.

for me relying on indicator or some random person calls is no different.

If you want to rely on something use Fibonacci. its useful in finding support and resistence levels. Fib is math, its more reliable than indicators.

1 Like

Yes, I am using moving averages and Fibonacci.

Can you explain what do you mean price action trading? I look at price and volume but can you explain.

Price action is simply the study of price movement in the market. you just have to understand how time and price works . every technical indicator is based on price. When seeing the live chart you need to be able to understand whats happening in the market and what is about to happen. In simple terms only based on price your trades will be.
Instead of making as a big story i’ll keep it simple.

combining both volume and price, patterns, trendlines, pullbacks, candle stick patterns, Breakout, Support and Resistence. All these you get from a plain simple chart, If your trades are based on these which you derive from chart, then you can call it as price action.

1 Like

Yes, doing all of this on a basic level.

For day trading, I find it very difficult to guess the next movies. I don’t have much experience of course, but still it is tough. For delivery trades, it is relatively easy.

doing intraday swing or positional doesn’t matter, its your cup of tea. but master price action, basic is not bad meanwhile it isn’t good too

Hi,

I strongly believe that manual Back-testing is the key to become a better trader because:

  1. It will give you a sence of “How market is actually working?” Like what levels market is respecting or rejecting?

  2. With Enough manual Back-testing, you will handle different situations better while trading.

  3. You will know when to take trades and when not to.

  4. As we all know that “Practice Makes a person a person perfect”. So why not take Manual Back-testing as a practice. You will learn more than theory.

You need to test for atlest one year, more is better. But the best way is to test your strategy for 3 scenarios:

  1. Up trend
  2. Down trend
  3. Consolidated market (Most Important)

Look at daily chart to get general idea for market scenarios. Then Back-test you strategy in lower Time frame.

Manual Back-testing is hard and time consuming, well that is the price you have to pay the market for your learning. (No shortcuts)

Lastly, I will say that you really don’t need to learn coding or use any software.

Just observe the price action. (Higher lows, Higher highs, lower lows, lower Highs, Support and resistance, Price and volume relationship).

Happy Back-Testing,

Dewansh Jaiswal

1 Like