Fully automated futures DAY trading

[Businessman]

Based on a 10 year backtesting sample on a basket of futures markets with my proprietary fully automated intra day trading system.

The monthly stats for the system are as follows:

• The monthly mean return is 5%.
• The monthly standard deviation is 10%.

I am going to assume the system's monthly returns are roughly Normally distributed.

As you can see from the image, the 3 standard deviation monthly range is -25% to +35%.
In the back test the monthly range was actually -17% to +33%.

All the above figures are based on a 1% bet size per trade.
If I could not handle the small (but real if i trade long enough) probability of a 20% losing month I could reduce the bet size to say 0.5% per trade.

In the back test 31% of months lost money (shown as black in the bell curve)
And 69% of month were profitable (shown as white in the bell curve)

Live trading this system will commence later today.
The backtest for the system is already up 1.6% for October 2020, and 32% so far for 2020.
2020 is turning out to be a sub par year for this system. But this is normal, 50% of years are sub par!
And so are 50% of months. The hope is I don't get many >10% losing months in a row!

The Sharpe Ratio for this system is around 1.7.
That might seem a bit high, but the system has limits on the number of contracts it can handle per trade.
It is going to struggle to scale to make 60% per year on $5 million for example.

I will update this journal at the end of each month. With monthly % PnL for the month based on live trading with a 1% bet size.

PS. I am not going to disclose any further information about this system (eg which markets it trades or how many signals it generates per day), sorry but I don't need any more competition.

 

[Dazz]

What platform system was it back tested on please? Goes to reliability of back testing itself.

 

[Scataphagos]

Based on a 10 year backtesting sample on a basket of futures markets with my proprietary fully automated intra day trading system.

The monthly stats for the system are as follows:

• The monthly mean return is 5%.
• The monthly standard deviation is 10%.

I am going to assume the system's monthly returns are roughly Normally distributed.

As you can see from the image, the 3 standard deviation monthly range is -25% to +35%.
In the back test the monthly range was actually -17% to +33%.

All the above figures are based on a 1% bet size per trade.
If I could not handle the small (but real if i trade long enough) probability of a 20% losing month I could reduce the bet size to say 0.5% per trade.

In the back test 31% of months lost money (shown as black in the bell curve)
And 69% of month were profitable (shown as white in the bell curve)

Live trading this system will commence later today.
The backtest for the system is already up 1.6% for October 2020, and 32% so far for 2020.
2020 is turning out to be a sub par year for this system. But this is normal, 50% of years are sub par!
And so are 50% of months. The hope is I don't get many >10% losing months in a row!

The Sharpe Ratio for this system is around 1.7.
That might seem a bit high, but the system has limits on the number of contracts it can handle per trade.
It is going to struggle to scale to make 60% per year on $5 million for example.

I will update this journal at the end of each month. With monthly % PnL for the month based on live trading with a 1% bet size.

PS. I am not going to disclose any further information about this system (eg which markets it trades or how many signals it generates per day), sorry but I don't need any more competition.

You've got it man... the Holy Grail!

1. Wake up
2. Turn on computer
3. Go to golf course
4. Come home, pick up the money, and take it to the bank.

Hakuna Matata!!

 

[DevBru]

You've got it man... the Holy Grail!

1. Wake up
2. Turn on computer
3. Go to golf course
4. Come home, pick up the money, and take it to the bank.

Hakuna Matata!!

Even better, run the algorithm 22 hours a day and wake up richer than you went sleeping last night.

 

[Businessman]

What platform system was it back tested on please? Goes to reliability of back testing itself.

I roll my own back testing software.

You've got it man... the Holy Grail!
4. Come home, pick up the money, and take it to the bank.

Sadly this system has 31% losing months, so is not the holy grail.
The Sharpe ratio, is 'just' 1.7. Not 3+, which would be the holy grail.
But yes in periods where it prints money it feels like the holy grail. eg When it is on a big winning streak.

Even better, run the algorithm 22 hours a day and wake up richer than you went sleeping last night.

I have to be selective with the trades i take and only trade RTH. Also if i could trade 22 hours per day the Sharpe ratio of this system would rise above 3.0 and then i really would have the holy grail. But sadly i have not found a system that good.

 

[Scataphagos]

Sadly this system has 31% losing month

Oh... perhaps you shouldn't use it all the time then.

Maybe just during golf season.

 

[Dazz]

Based on a 10 year backtesting sample on a basket of futures markets with my proprietary fully automated intra day trading system.

The monthly stats for the system are as follows:

• The monthly mean return is 5%.
• The monthly standard deviation is 10%.

I am going to assume the system's monthly returns are roughly Normally distributed.

As you can see from the image, the 3 standard deviation monthly range is -25% to +35%.
In the back test the monthly range was actually -17% to +33%.

All the above figures are based on a 1% bet size per trade.
If I could not handle the small (but real if i trade long enough) probability of a 20% losing month I could reduce the bet size to say 0.5% per trade.

In the back test 31% of months lost money (shown as black in the bell curve)
And 69% of month were profitable (shown as white in the bell curve)

Live trading this system will commence later today.
The backtest for the system is already up 1.6% for October 2020, and 32% so far for 2020.
2020 is turning out to be a sub par year for this system. But this is normal, 50% of years are sub par!
And so are 50% of months. The hope is I don't get many >10% losing months in a row!

The Sharpe Ratio for this system is around 1.7.
That might seem a bit high, but the system has limits on the number of contracts it can handle per trade.
It is going to struggle to scale to make 60% per year on $5 million for example.

I will update this journal at the end of each month. With monthly % PnL for the month based on live trading with a 1% bet size.

PS. I am not going to disclose any further information about this system (eg which markets it trades or how many signals it generates per day), sorry but I don't need any more competition.

So let me:
you have 10 years of back test data done on a system we don't know, cannot see and do not understand and now having never taken a single trade, you suddenly burst onto the stage, all with zero performance, to tell us you are now going to trade it? Call me underwhelmed - why not have not waited to next year instead?

 

[Businessman]

Call me underwhelmed - why not have not waited to next year instead?

Im hoping for some General Election volatility, so a good time to start. Also the period November through January, seem to be above average for the system.
This could just be random. But i think there is more money moving around at the end of the year and the start of the year. Compared to the middle of the year.
Of course not every November to Jan period shows this out performance. There have been years when these months have been lousy for the system. Again this could just be random and a result of the small sample size (number of years).

 

[Dazz]

Im hoping for some election volatility, so a good time to start. Also the period November through January, seem to be above average for the system.
This could just be random. But i think there is more money moving around at the end of the year and the start of the year. Compared to the middle of the year.
Of course not every November to Jan period shows this out performance. There have been years when these months have been lousy for the system. Again this could just be random and a result of the small sample size (number of years).

Call me underwhelmed - why not have not waited to next year instead to show us 1 year trade data? Who cares about back test data on a system we have never seen or heard of? Striker has back test data on 2944 trade algo systems, iCannon an other 1900.

 

[Businessman]

Call me underwhelmed - why not have not waited to next year instead to show us 1 year trade data? Who cares about back test data on a system we have never seen or heard of? Striker has back test data on 2944 trade algo systems, iCannon an other 1900.

Well i am now, as of today, trading the system live. It is no longer just a theoretical back test. Plus this Journal is for my own benefit.

This Journal is deliberately designed to be underwhelming and boring, so i can concentrate on the process of trading and not the results. I wont be talking about $dollar figures made or lost. That is exiting and entertaining i know. But distracts from the process of trading. Which is to follow the system correctly through good times and bad.