HFT Impact on a Futures Retail Trader

[i960]

how can HFT algos screw me?

By running that 3 tick stop enough times that your risk/reward ratio results in negative expectancy.

The strategy would be successful if you're accurate enough trading with momentum and/or pullbacks etc by you've surely got to be trading with the algos rather than against them here.

You should also examine how *time* efficient it is as well. If you're basically locked to a chair all day to make 50 ticks total (this assumes you're winning every single day too) that might not be the best use of your time unless you're trading 10+ lots.

 

[CyJackX]

My belief is that, unless you operate on HFT timeframes, you shouldn't care about HFTs at all.
Worrying about them is sheer paranoia in these contexts.

"They'll run my stops!"
Who can truly be blamed for running your stops? You really think they're going to put that much risk on the line moving the market to move you out of a trade?

Even if you rationalize it as the HFTs blowing thru the "common, easy" stops to nab a group, that's still overly simplistic.
Any "algo strategy" that is simple enough for you and I to conceptualize has already had its edge traded away. There are multiple HFTs, and they can only share the pie so many ways until the pie doesn't exist. HFT strategies operate on algorithms that are proprietary and researched and in no way as simplistic as imagined by tinfoil hat theorists here.

 

[algofy]

HFT is going to be rushing for queue position when a level is taken out and also trying to play the inside bid/ask game which you aren't playing. So I don't think it's going to mess with you too much. The bigger concern is the negative risk/reward like someone else stated. You have to have a pretty high win rate just to break even. Plus commissions are going to eat pretty heavy into your profits. What instrument are you trading? Since you said ticks I assume futures. Risking 3 ticks in CL is a much different game than risking 3 ticks in the 2 year tsy futures.

 

[cjbuckley4]

HFT is going to be rushing for queue position when a level is taken out and also trying to play the inside bid/ask game which you aren't playing. So I don't think it's going to mess with you too much. The bigger concern is the negative risk/reward like someone else stated. You have to have a pretty high win rate just to break even. Plus commissions are going to eat pretty heavy into your profits. What instrument are you trading? Since you said ticks I assume futures. Risking 3 ticks in CL is a much different game than risking 3 ticks in the 2 year tsy futures.

If all the HFT players are interested in the good Q leaving him the bad, what will happen when he's the only one left in the Q?

Adverse selection. Sophisticated traders will see that taking out the remaining contracts at a given price level is similar to having #1 Q position on the inside of the newly created price level. Thus, if he's happy to occupy the end of the Q he should know that he's going to often be starting 1 tick in the hole.

 

[eS19]

Thanks all for your helpful comments. This is my third week trading ES. I only trade ES. I agree with most posters here, and I can attest to the fact that the way HFT impact my trading is by beating me to fill of my order. We are all looking at the same market and the way they beat non HFT traders is how fast can they get fill when they know an impending move. Usually 1-2 ticks.

Thanks again,
Sid

 

[algofy]

Thanks all for your helpful comments. This is my third week trading ES. I only trade ES. I agree with most posters here, and I can attest to the fact that the way HFT impact my trading is by beating me to fill of my order. We are all looking at the same market and the way they beat non HFT traders is how fast can they get fill when they know an impending move. Usually 1-2 ticks.

Thanks again,
Sid

I don't think HFT *knows* about an impending move, they may rush in to get an early bid/offer once a price levels clears out and offer/bid the other side immediately but several ticks away from the inside market where you are located, I don't think HFT is *beating* you. They already played their game at that level several ticks ago. They have the speed to play the inside bid/ask game and also have the commission/exchange fee structure to do it. If they can get first to a bid offer and get filled very quickly, and the momentum starts to die, they can simply market out for a scratch and their costs are ridiculously cheaper than what you and I pay. So when they scratch they might lose 20 cents but when the momentum continues and they pick up a tick, they make 12.30. It's a pretty good risk reward scenario and it's only possible because of their speed and pricing structure.

For a point and click retail trader (I'm assuming this is you) dropping limit orders in a few ticks away from the current inside market, I really don't think HFT is affecting you too much, at least in how orders are getting filled.

As said previously, were I you, I would be much more concerned with your wrongly skewed risk/reward profile. maintaining a 70%+ win rate consistently is very difficult and low profits mean commissions and exchange fees really play a big role. So in your example and assuming a $4 RT on ES fees:

Winning trades (2 ticks) net you $21
Losing trades (3 ticks) cost you $41.50

so in order to BREAK EVEN you need just about 70% winners.

I'm not saying it's impossible on a short time frame to get 70% winners but I know I've tried that math and I can't make it work on my end. Finally, if you're going to do this strategy, you really need to have the lowest costs possible, if you are paying TDAmeritrade prices, forget about it. You need to be at least $4 or lower and don't let anyone fool you, with a strategy like this, a nickel or a quarter lower can make a big difference because I don't think it makes a ton of sense on a strategy like this to take 2 or 3 trades a day. This is a manual high frequency model. All of this IMO of course. Good luck.

 

[algofy]

^ Followup

I might considering doing some testing and playing around with the math some. Are you looking at charts or using the depth of market? If using charts, you may be able to look back and see how often your 2 tick moves would have yielded 4 ticks. If you can get your average winner to be 4 ticks and maintain your average loser at 3 ticks. Now you can have a 50% win rate and if consistent can goose the strategy with some size and make a pretty decent amount.

 

[eS19]

Thanks. If I am buying, i am putting my limit one tick below last bid. If i am selling, my limit is one tick above last ask. I am timing market at key s/r levels. I use ATM strategy to target 2 ticks on half position and 3 ticks on the other half while moving the stop to breakeven-1 when first target is hit.

I have had poor results when I tried this with moving the stop to BE + 0. I will almost always get stopped out on that second half. I have not tested with target 1= 3 ticks and target 2 = 4 ticks.

Regarding your earlier post, I did not get a fill on 4-5 orders today. They were almost at the bid/ask. And the price took off the moment I put my limit orders. I do not think I will get a fill at all 2 ticks below last traded price, given the way my strategy works.

 

[algofy]

Thanks. If I am buying, i am putting my limit one tick below last bid. If i am selling, my limit is one tick above last ask. I am timing market at key s/r levels. I use ATM strategy to target 2 ticks on half position and 3 ticks on the other half while moving the stop to breakeven-1 when first target is hit.

I have had poor results when I tried this with moving the stop to BE + 0. I will almost always get stopped out on that second half. I have not tested with target 1= 3 ticks and target 2 = 4 ticks.

Regarding your earlier post, I did not get a fill on 4-5 orders today. They were almost at the bid/ask. And the price took off the moment I put my limit orders. I do not think I will get a fill at all 2 ticks below last traded price, given the way my strategy works.

I hope this works out well for you. Is it automated or manual ?

 

[birzos]

Limited if any, only if you put your trades around large lots, and impacted if your data feed includes HFTs.

The first will impact negatively and is pot luck, the second you may gain, be neutral, or lose, depends on strategy.

There are bigger problems to worry about, it is rare anyone makes any profit at that level without negative return vs effort.