My algo trading daily performance

[alroyraj]

Hi nj23,

It really depends on the trading day. Sometimes there are 2-3 trades, Sometimes there are 20-22. On an average is 12-13 each day. The maximum being 43 in the backtest.

Thanks
MT

I think the importance of trading capital comes in here as in a trending system when there is a big trending move there needs to be more simultaneously long positions. Hence I believe mechtrader bhai you rotate the watch list so as to catch the big move as per the afl.

 

[alroyraj]

Interested in knowing how mechtrader bhai 's alho trading survived the Monday crash or if you needed to build in max loss per day.

 

[mechtrader]

Interested in knowing how mechtrader bhai 's alho trading survived the Monday crash or if you needed to build in max loss per day.

Monday was all short for me.. It was surely a day to remember.

Over the past two months, I had been noticing that I was loosing MTM profits during reversals as I use very lenient TSL. In over 4-5 occasions the MTM profit of 50k was reduced to 10k or less.
This month I decided I will book all positions at 50k MTM. Did the same 3 times before monday. On Monday also did the same. I booked at 12:50 pm. Little did i know that it would be the biggest day of my trading career. The model reached 6 times the profit booked.

Though my approach was correct, but it was not tested. Maybe if i test this scnario, that book profits at a particular value and after the testing the results are pretty much similar then my decision on monday was correct.
BUt the mistake that i committed is i implemented some thing in market without testing it. It was and still is a huge blow. Trying to cope up and not making the same mistake again.

P.S. Sorry for such a long write up. Its the emotions that re speaking.

Thanks,
MT

 

[alroyraj]

Monday was all short for me.. It was surely a day to remember.

Over the past two months, I had been noticing that I was loosing MTM profits during reversals as I use very lenient TSL. In over 4-5 occasions the MTM profit of 50k was reduced to 10k or less.
This month I decided I will book all positions at 50k MTM. Did the same 3 times before monday. On Monday also did the same. I booked at 12:50 pm. Little did i know that it would be the biggest day of my trading career. The model reached 6 times the profit booked.

Though my approach was correct, but it was not tested. Maybe if i test this scnario, that book profits at a particular value and after the testing the results are pretty much similar then my decision on monday was correct.
BUt the mistake that i committed is i implemented some thing in market without testing it. It was and still is a huge blow. Trying to cope up and not making the same mistake again.

P.S. Sorry for such a long write up. Its the emotions that re speaking.

Thanks,
MT

Excellent to see your model worked this Monday. It's always a trade off between booking out and trailing the profit. These instances come very rarely that we would not think of testing these types of movements.
Since you use some sort of candlestick patterns you captured the movement well. Not very often do you capture the entire move. Look at indiabulls real estate it crashed 12% which is around 6 points and bounced 20% which is around 9 points.
These next month or so should be a time for consolidating and your current approach would yield better results.
Even theoretically when building a model it is advised to test without these large range profit days so we can test the consistency of the model.
When trading these discretionary we have the option of closing one position and trailing the other but in an algorithms this is a decision you have to make earlier.

 

[mechtrader]

Excellent to see your model worked this Monday. It's always a trade off between booking out and trailing the profit. These instances come very rarely that we would not think of testing these types of movements.
Since you use some sort of candlestick patterns you captured the movement well. Not very often do you capture the entire move. Look at indiabulls real estate it crashed 12% which is around 6 points and bounced 20% which is around 9 points.
These next month or so should be a time for consolidating and your current approach would yield better results.
Even theoretically when building a model it is advised to test without these large range profit days so we can test the consistency of the model.
When trading these discretionary we have the option of closing one position and trailing the other but in an algorithms this is a decision you have to make earlier.

Very much valid points made by you. Your understanding is good. I desperately want to perform this testing. But its hard work. Hope all works out.:thumb::thumb:

Till the time testing is not done, i will stay with the current approach.

Thanks,
MT

 

[mechtrader]

Trading performance for Aug series. It was a brilliant trading month, but i lost out on opportunities due to some other rules. Nevertheless, the mont was quite good.:thumb:

Thanks,
MT

 

[mithoon]

1. Yes the first trade starts after 9:25.
2. all trades are strictly intraday.
3. Yes, all in stocks futures.
4. I have tried various things for stop losses, atr based, some ma, but i found that nothing works better than a fixed percent stop loss. I use 1-2% as the stop loss depending on the volatility in the market. It is % of position taken.
5. I have designed all strategies as trend riders. I dont use any profit targets. So, the win percentage is low. I have around 40% winners.

These are stats for one of the strategies.

The stocks list is:

Thanks,
MT

Hi MT,
Isn't Stop Loss based on percentage Position Size the same as Stop Loss based on percentage from close price
Example
Buy price = 10, Position size = 10,000
Shares = Position size / Buy price = 1,000

Lets say price decrease by 10%:
Close price = 90% X 10 = 9
Position size at price = 9 is Shares X price = 1,000 X 9 = 9,000
So you see that if price decrease by 10% (10 to 9), position size decrease by 10% (10,000 to 9,000)

 

[plyfz]

Pls post a video too which would validate

 

[mechtrader]

Hi,

Sept series was a good one to trade.:thumb::thumb:

Thanks,
MT

 

[nifty trade]

@mechtrader

Can you share good way to find smoothly trending scrips ?