Index futures automation

[NoahA]

I hope my algo doesnt break when I switch over to NQ. A lot of it depends on high volume (low slippage)

I wanted to ask you about this because I see things a bit differently. You seem to call "lack of slippage" as high volume, but to me, MNQ, even with the higher volume, has lots of slippage because its supposed to track the NQ, but often during times of spikes, can easily deviate by several points, and hence 10 or 20 ticks.

Now I'm not sure if your algo needs this, but this is what I see. If NQ for example pokes a key level by only 1 point on a spike up, the MNQ could poke several points. So if you had a stop there lets say, the fill could be much worse, depending on which direction you are playing of course.

But ya, even with the higher volume for MNQ that you showed, I think slippage is much worse.

 

[Businessman]

I wanted to ask you about this because I see things a bit differently. You seem to call "lack of slippage" as high volume, but to me, MNQ, even with the higher volume, has lots of slippage because its supposed to track the NQ, but often during times of spikes, can easily deviate by several points, and hence 10 or 20 ticks.

10 or 20 tick deviation in the price of MNQ relative to NQ is not necessarily the same as 20 point slippage on your MNQ fills.
As long as you have backtested on MNQ data and your system works, those deviations from NQ should not be a problem as long as your fills are still ok.
But if you are often getting large slippage on your fills during regular trading hours (not counting CPI/FOMC spikes etc), then it could mean live trading is nowhere near as good as the backtesting suggests.

 

[Overnight]

I wanted to ask you about this because I see things a bit differently. You seem to call "lack of slippage" as high volume, but to me, MNQ, even with the higher volume, has lots of slippage because its supposed to track the NQ, but often during times of spikes, can easily deviate by several points, and hence 10 or 20 ticks...

I have to think that is just an illusion of the last-trade price display, as a 10-20 tick difference would be quite an arb opportunity that I am sure was exploited to meaninglessness back in May 2019.

Any inefficiencies in a mini-micro spread, if they still exist, must be getting filled in microsecons by the algos.

 

[hilmy83]

Supposed to be a B/E day. The EA was supposed to cover the cost of commission, but it didn't quite do so. I suspect the larger position size created larger slippage. But everything did actually execute as intended.

View attachment 324009

View attachment 324010

 

[hilmy83]

Monthly performance:

Aug: +$1247

Took out the outlier day pnl.

View attachment 322254

Monthly performance:

Sep: +$1568.5

Took out the $2458 outlier profit due to QT algo malfunctioning on 9/11.

If I was single and living in a studio...I could live off this

 

[hilmy83]

View attachment 324108

View attachment 324109

 

[hilmy83]

View attachment 324355

View attachment 324356

 

[SimpleMeLike]

View attachment 324434

View attachment 324435

Good Morning hilmy83,

Just found your thread. Awesome work buddy. I am proud of you and appreciate you for showing your trading results to the public. I will do same once I get my algo lined out. Still fine tuning algo for NQ or ES. I prefer ES, but algo likes NQ. Two algos I had running first of the year experienced brutal drawdown, looking for something with less drawdown.

Enjoy your day.

 

[hilmy83]

Good Morning hilmy83,

Just found your thread. Awesome work buddy. I am proud of you and appreciate you for showing your trading results to the public. I will do same once I get my algo lined out. Still fine tuning algo for NQ or ES. I prefer ES, but algo likes NQ. Two algos I had running first of the year experienced brutal drawdown, looking for something with less drawdown.

Enjoy your day.

Follow the money man, don't dwell on a market just because you like it. If NQ shows it's profitable, go with that.

 

[SimpleMeLike]

Follow the money man, don't dwell on a market just because you like it. If NQ shows it's profitable, go with that.

Hello hilmy83,

Thank you for that advice buddy. Appreciate it and you right.

How many algos are you running?