HFT is Killing the EMini

Quote from syswizard:

Exactly. Most of the complaints about HFT come from those who can't program and don't have a clue about how an algo is constructed and tested.
More...

Spot on.

Not only that, no clue as to how HFT traders get paid. They are paid for filled resting orders. That is they make mean reversion trades which dampens noise and adds great liquidity.

Daily range is up, average volume is up and volatility is up - doesn't sound like a dying anything to me - more like a traders dream.

I trade order flow with spikes that detect surges in commercial trade velocity and have seldom seen such a favorable enviornment for short term, intra-session trend following trades.

Big legs, more volume and someone is bitching? If HFT is the cause of this - I say bring more of it.

Attached is a chart from Friday's trade (09/09 Time PST) with just such a situation. There was a topping spike in commercial trade velocity and obvious negative divergence between price and a moving window of order flow.

In the 40 monutes following the spike and the divergence ES dropped 20+ points. Not bad for a market that is being killed by HFT.
 

Attachments

Quote from HTmarket:

If you're a trader and you've made some money. Why not do a little research and find a strategy that can be programed. Read a few programming books, and work with some smart computer science kid to help you code it. You will both learn, you about programming and him about trading.
More...

:confused: :confused: :confused:
 
Quote from austinp:

The difference in ES now versus years past (I have traded SPX options or ES since late 1999) is the structure of price movement.

Before the directional turns or transitions were mostly rounded, cup & handle, 1-2-3 type sequences. That came from thicker open interest, two-way markets and deliberate accumulation => distribution behavior.

Now the general price structure is all sudden v-turns, and most true turns follow higher high or lower low swings before a sharp reversal slams the other way.

So it was easier in the past to add size on positions as the price action stepped higher or lower, and it was easier to trust what you saw because of more deliberate price action.

Now the algos strip several strikes or handles going one way, take it a few ticks beyond to trigger resting stops and then rip back the opposite direction to clear that side, too.

ES traders never really know if they are seeing a pullback against the "trend" or a complete out-of-blue reversal that's running deep the other way.

**
More...



This is precisely whats going on.
 
--------------------------------------------------------------------------------
Quote from syswizard:

Exactly. Most of the complaints about HFT come from those who can't program and don't have a clue about how an algo is constructed and tested.
--------------------------------------------------------------------------------
Greetings,

I don't have much of a clue as to what algos do, or how they are constructed. And, I do very much agree with Spiker and Austinp on their market views on algos. This conclusion is based on my market experience and what I have witnessed on the charts through more screen hours then I care to mention.

However, since you appear to have some expert insight as to how an algo is "constructed and tested". And, I'd be very much interested in learning your opinions on these topic areas, or any other insights you have. I would like to learn and know "your specific" conclusions as to how algos are constructed and tested, and not some book definition or reference.

I would like to learn, what a person who has mastered these concepts knows, at least in general. Could you please share some of your insights on these two topic areas. My basic questions are:

1. What specifically are these algos constructed to do?
2. What specific conditions are the algos looking for in the market?
3. What procedures do they use to accomplish this task?
4. How do they test for efficacy?
5. In your opinion, how would one go about protecting themselves from an algo strike?
6. In your opinion, how would one go about taking advantage of a potential algo strike?

I would like to learn how to better coexist with these algos, since they appear to be here to stay......at least for the foreseeable future.

Big Thanks!
 
Quote from SrRuthenate:

That 9000 HFT order is not a normal limit order. It's probably a FOK (fill or kill) order or a variation thereof. There are ways to identify this and exploit it so that you can FOK them and they can't FOK you.:eek:

Nothing mysterious either. That 9000 bid you see is really a series of intermittently sent (fast) series of FOKs that also have a latency and a series of rules. When you find out the latency of the algo and a faint idea of the rules it runs under, you wait till it zigs and then you FOK it. In other words calculate when your bullet order is going to arrive at the exchange and syncro it with the zig of the faster algo. Then mark the market quicly against it till it pukes. Simple really.
More...

Shouldnt the second FOK order be more faster than the first one?
 
While I understand now HFT take advantage of collocation to trade equities, and make money from rebates, sub penning, flash orders and etc, I still find it hard to understand how these strategies can be transferred to ES.

If anyone can shed any light, that would be great.

Thanks
 
Whats ruined the mini's is the bad economy and bad Fed policy coupled with the wipeout of many market participants two years ago. HFT would have been great in 08, but look at the returns today. They arent that great. They need retail and institutions to make a living and they are gone.
 
Quote from rosy2:

everyone has that priviledge.



what makes you think they are the fastest? :confused:
More...

yeah, there is a relative news blackout [as much as there can be one in this internet age] about quantum computing.. I thought it was about "National Security" but maybe it's about "Hedge Fund Security"...
 
Quote from T1P1:

Spot on.

I trade order flow with spikes that detect surges in commercial trade velocity and have seldom seen such a favorable enviornment for short term, intra-session trend following trades.

In the 40 monutes following the spike and the divergence ES dropped 20+ points. Not bad for a market that is being killed by HFT.
More...
Very nice analytics you've got going on here.
Trading on volume bars of 1000 contracts, eh ?
Very interesting.
 
Quote from KDASFTG:


However, since you appear to have some expert insight as to how an algo is "constructed and tested". And, I'd be very much interested in learning your opinions on these topic areas, or any other insights you have. I would like to learn and know "your specific" conclusions as to how algos are constructed and tested, and not some book definition or reference.

I would like to learn, what a person who has mastered these concepts knows, at least in general. Could you please share some of your insights on these two topic areas. My basic questions are:

1. What specifically are these algos constructed to do?
2. What specific conditions are the algos looking for in the market?
3. What procedures do they use to accomplish this task?
4. How do they test for efficacy?
5. In your opinion, how would one go about protecting themselves from an algo strike?
6. In your opinion, how would one go about taking advantage of a potential algo strike?

I would like to learn how to better coexist with these algos, since they appear to be here to stay......at least for the foreseeable future.

Big Thanks!
More...
Dude - I buy all of the HFT books and COMBINED they wouldn't answer ANY of these.....not that they couldn't !
Everything is "hush, hush" in the HFT trading arena.
 
You must log in or register to reply here.
Back
Top