Jump to content

Welcome to the new Traders Laboratory! Please bear with us as we finish the migration over the next few days. If you find any issues, want to leave feedback, get in touch with us, or offer suggestions please post to the Support forum here.

  • Welcome Guests

    Welcome. You are currently viewing the forum as a guest which does not give you access to all the great features at Traders Laboratory such as interacting with members, access to all forums, downloading attachments, and eligibility to win free giveaways. Registration is fast, simple and absolutely free. Create a FREE Traders Laboratory account here.

Recommended Posts

I have actually been working on algorithmic scalping recently and just started live trading the strategy this week. From what I can gather, it can't really be done efficiently by hand anymore. Slow reaction time not only makes you miss out on a lot of opportunities, but also also dramatically increases your risk.

 

However, doing it algorithmically is also difficult. The code to find the right market conditions to begin scalping is non-trivial and your code needs to be smart enough to stop scalping when conditions change (especially if you are doing it on a large scale over many symbols). However, if this is done properly (I think my implementation is fairly robust now after much trial and error), scalping IS still possible.

 

The question now is whether it is scalable. Small size non-hidden limit orders sent directly to the NASDAQ matching engine via OUCH has a decent chance of getting filled, but with so many quant scalpers out there nowadays, it's unclear whether it is possible to get a sizable enough number of shares to really roll in the big bucks.

 

Hi Swing_Trader

 

Unless you want to be dealing with the kind of sums that the big HFT firms work with, surely as a solo trader you could make a decent income without needing to scale?

 

Have results since you began live trading broadly matched your expectations?

 

BlueHorseshoe

Edited by BlueHorseshoe

Share this post


Link to post
Share on other sites
search for posts containing "algo,bot" by steve46 in "Day Trading the E-mini Futures" returned 6 results:-

 

search query results

 

Hi Neg,

 

Thanks for doing the leg work.

I have looked at the Steve's six examples and yes he has highlighted a recurring

pattern of behaviour where price does a test and then retest of a level followed by a retrace.

Usually, as has been mentioned this zone lies to the outer edge of an s&r zone or a zone of little activity.

And yes it is a good setup for a retrace.

But this setup has been around for years in my memory and perhaps bots have picked up

for this very reason.

This conversation concerning bots is a good conversation IMO, but it needs to focus around exactly what bots are doing to the market that Traders are not already doing.

 

I remember the conversations concerning the use of Indicators and it became apparent

that the latest fashion was to trade without them.

 

As MM has just pointed out ... the only way to trade is profitably

Or as my Broker points out ... it doesn't matter what you do so long as you make money.

 

ST makes a good point that bots could effect order entry and I would be interested to hear from someone who knows how solid the order enter queue really is for the CME [as an example]

 

And so, I like some others here am looking for substance in the matter of bots and their effect on the markets...

I know very little about the subject of bots because I am completely removed from the People who control and build them, and that is why this conversation is important to me.

 

And because it is important it would be a shame to see it disintegrate into a fashion item because bots are the new new thingy.

 

Thanks again Neg for your legwork and thanks Steve for posting the six charts .. I know it all takes valuable time.

Edited by johnw

Share this post


Link to post
Share on other sites
Thanks for that remarkably edifying contribution Steve - we were trying to discuss scalping and HFT though, not canine grooming habits.

 

If you want to chat more about your neighbour's dog, then I would suggest one of the following outlets:

 

Dog Forum at Dogsey - the worlds largest dog forum!

Dog Forums - I-Love-Dogs - your Ultimate Dog Forum

Barking Mad - forum for dog lovers. - Portal

 

Best of luck with that.

 

BlueHorseshoe

LOL S46 won't be back........for a while. He tends to get mad at what he thinks is inferior intelligence and then tramps off for a while...but he will eventually come back....always does. Scalping of course can be done 2 ticks to 1 pt is about what I consider a scalp. This can be done in spite of bots..algo..hft...etc. There is no way to figure out all the bots....even if one could, there is no way to effectively use the info for trading purpose. a new bot would come along. then 100...then 1000...then 10,000. My ideas for scalping is scalp with trend using pullbacks. Scalp with trend in strong trends. Scalp with good reversals. Scalp trading ranges by shorting at top and long at the bottom.

 

Bots or no bots there will always be trends..pullbacks...trading ranges. Always have existed and always will. It is the nature or DNA of the markets. Regardless of who makes the trends...pullbacks..ranges..bots..hft.. or people. So my theory is take what always happens and trade it. If ones timing is wrong and their stops keep getting hit then they could be right strategically but using wrong tactics and will need adjustment to their implementation of the strategy. Even the much vaunted reward to risk calculation may need some modification for scalping purposes in the world of bots...hft..etc.

Share this post


Link to post
Share on other sites

Hello Pattuca

 

For the record I am (I have said this once before in this thread) teaching a class so the only time I can visit is during coffee breaks.

 

and once again for the record, I don't view the world in the way you propose. "inferior" isn't part of my vocabulary....I do however think of folks as either adult or infantile....and since I have already raised children, I am not interested in putting up with infantile behavior here..

 

JohnW

 

Thanks for taking the time to look up the data....you took the time and made the effort to evaluate the information like an adult...all thats missing is the context...I will talk about that in my thread on "characterizing markets". Hopefully I can show folks how to integrate the concepts of Market Profile with automated execution so that they use it to improve their trading.

 

Good luck

Share this post


Link to post
Share on other sites
LOL S46 won't be back........for a while. He tends to get mad at what he thinks is inferior intelligence and then tramps off for a while...but he will eventually come back....always does.

 

Just to be ABSOLUTELY clear - the pattern Steve46 identifies is in my opinion valid and perfectly tradeable. If this is the sort of strategy he teaches his students then they could certainly do far worse with other vendors.

 

However, I personally find it unlikely that this type of movement is a source of profit for HFT firms. Steve talks about movements of twelve to sixteen ticks; the average HFT position is reported as lasting approximately 22 seconds. Not many sixteen tick movements occur in 22 seconds.

 

Even supposing that this pattern is precipitated by the accumulative actions of HFTs, this offers no real insight into how such algos operate for single tick profits.

 

BlueHorseshoe

Share this post


Link to post
Share on other sites

Gentlemen, enough already! As much as we are all are entitled to our own opinions, there is a line. Disagreement and debate is fine. In fact, it is good so long as we consider each others' ideas rather than just try to 'win' the debate. Thus far, I'm pretty interested in what has been said. I can't remember where I saw it but for those interested in HFT etc, it might be worth searching for the recent DTN webinar on youtube (hosted by BMT I believe).

Share this post


Link to post
Share on other sites
Just to be ABSOLUTELY clear - the pattern Steve46 identifies is in my opinion valid and perfectly tradeable. If this is the sort of strategy he teaches his students then they could certainly do far worse with other vendors.

However, I personally find it unlikely that this type of movement is a source of profit for HFT firms. Steve talks about movements of twelve to sixteen ticks; the average HFT position is reported as lasting approximately 22 seconds. Not many sixteen tick movements occur in 22 seconds.

Even supposing that this pattern is precipitated by the accumulative actions of HFTs, this offers no real insight into how such algos operate for single tick profits.

BlueHorseshoe

 

 

I can't answer for Steve, but this pattern is created by s&r zones as has been mentioned, and occurs several times per session on ES.

It is highly accurate and can run 16 tics quite quickly.

 

The run size and speed depend on two events.

1.. if another s&r zone lies just outside the zone that this pattern is testing then pressure is been added to the Trader to fade the pattern.

2 .. if a third zone lies ahead of the intended fade then the price will run to the this zone almost without doubt and the distance to this 3rd zone can be 12-16 tics.

Granted, the faded trade must push through the original zone that created the pattern, but it's appetite to cause trouble has been largely eroded by the pattern itself as it chewed through stops on the fade side of the congestion.

The premise to keep in mind at all times is that price waves start at the opposite sides of congestion before eventually breaking out.

 

As for 22 seconds, I am here to learn about HFT so I cannot comment.

But I imagine HFT trades have a long tail distribution and a breakout trade from the pattern we are discussing lies in that long tail or as an outlier.

But does it really matter?

If I kept an analytical log of trades (which I don't) or if I kept a diary (which I don't), I would find that my distribution of my trades was asymmetrical as well, and that is the outcome that I want and work towards.

Share this post


Link to post
Share on other sites

Taking advantage of the small time lage between the market as a whole and an individual market maker. Commonly known as laitiency trading. Though probably mis-spelled.

 

It is less of an issue these days as most market makers are updating prices many times faster and quite a bit more accurately than was the case a few years ago.

 

Any time you trade with a market maker and treat them as an opponent rather than a clearing house you should expect them to figure out your posture and send you packing.

Share this post


Link to post
Share on other sites

Massive scalping can also be used as a form of "hedging". To create a "cushion" to protect larger trades.

 

See for instance this thread:

 

Forums - Algorithmic Trading a large folio with small orders

 

or this post on Linkedin:

 

I would be curious to share experiences: For you, what are 1- the most efficient hedging strategy you use? 2- the most efficient way to reduce drawdown? Thank... | LinkedIn

Share this post


Link to post
Share on other sites

John W - you should read up on Irene Aldridge if you want to learn more about HFT.

 

Personally, I think it's a waste of time unless you really want to use HFT as your strategy.

 

If you're the average retail day trader, your efforts will be better off learning about swing traders or who ever is up on the next time frame. The best opportunities are usually found by taking short term advantage of longer term opportunity. Therefore, looking at a smaller time frame down stream may seem wise and comendable, but I think it's a fools errand (JMO)

 

Of course, some vendors clearly like to jump in and pretend they have figured it all out in order to shill in the next batch of customers. Fear, uncertainty and doubt. These are the shills main tools. What better phantom than those nasty HFT geeks who can surely out smart you and send you to the poor house? The fact that many HFT trade passively as well as aggressively means they also help you get filled. The shills wont tell you that though.

 

You can spot the shill activity as they tell you that HFT is all momentum based stuff. In fact, momentum strategies are the highest risk strategy of all, so the least practised strategy in any HFT. Whats more, HFT firms will have MANY different algorithms/methods employed. Momentum being one (a small one). I'll tell you that for free so you dont have to pay a shill to tell you that!

Edited by TheDude

Share this post


Link to post
Share on other sites

I don't think so that scalping is dead. Indeed, If you want to make profits fast and like action, scalping is for you. Scalping is about getting a little profit on little trades. Many small trades a day gets a big profit at the end of the day. Be a careful pro. Scalping is the FX trading style for currency traders who want to make profits on a daily base. 1-2% profit every day, sounds good? That’s what professional scalper makes.

 

The main benefits from scalping:

 

High win rate

You can make high profits faster than with any other trading styles

:crap:

Share this post


Link to post
Share on other sites

Here is one reason why HFT has made scalping harder:

 

Note that scalpers use limit orders (except on stops). Here are 2 situations:

 

1. assume a scalper has a bid in at 1434 waiting for a retracement. His bid will be sitting behind a large number of HFT orders. If 1434 is a price of strong support, all the HFT orders must be hit before his order is. If he does get hit, and puts in his sell at 1435, his order again will be behind all the HFT orders, again making it less likely that he will get hit than before the "HFT era."

 

2. assume, instead, that the market is moving down and 1434 is no longer a place of support. (The algos will determine this in a very small fraction of a second). As soon as the first market sell at 1434 is made, all the HFT bid orders will disappear, and the scalper will be immediately filled in a declining market.

 

Result: Fills are worse now than in the pre-HFT era. Therefore: the more important getting good fills is, the less likely success with the trading method has become. For swing trading small orders, the difference is very small. For scalping ticks, it is quite significant.

Share this post


Link to post
Share on other sites

I think it's dead at dd brokers (because of their greed) but if you find a good stp/ecn broker with low latency technology and good conditions, I think scalping is still one of most popular trading strategy.

 

Check our equinix brokers for fastest order execution and market deepth I found it best for high frequency trading

Share this post


Link to post
Share on other sites

I used to trade some time ago (and I still do to some extend) the eurusd pair for short to very short targets (10 pips or even 5 pips) after identifying a range......it was a good system in the sense that when markets did not move, with 10 profitable days you would have had a 100 pips day on a non moving day.......and trade it without a stop loss

 

the problem was when on the wrong side of the market, is extremely difficult to exit.....

Share this post


Link to post
Share on other sites
In general, scalping is dangerous.

 

In general, trading is dangerous. Maybe you can be more specific?

 

The majority of my trades are in the 8-15 pip win size. I don't know if you consider that scalping or just short-term trading. For me, it is the easiest way to trade -- I've tried a large variety of approaches over the last 11 years.

 

With kind regards,

MK

Share this post


Link to post
Share on other sites
Here is one reason why HFT has made scalping harder:

 

Note that scalpers use limit orders (except on stops). Here are 2 situations:

 

1. assume a scalper has a bid in at 1434 waiting for a retracement. His bid will be sitting behind a large number of HFT orders. If 1434 is a price of strong support, all the HFT orders must be hit before his order is. If he does get hit, and puts in his sell at 1435, his order again will be behind all the HFT orders, again making it less likely that he will get hit than before the "HFT era."

 

2. assume, instead, that the market is moving down and 1434 is no longer a place of support. (The algos will determine this in a very small fraction of a second). As soon as the first market sell at 1434 is made, all the HFT bid orders will disappear, and the scalper will be immediately filled in a declining market.

 

Result: Fills are worse now than in the pre-HFT era. Therefore: the more important getting good fills is, the less likely success with the trading method has become. For swing trading small orders, the difference is very small. For scalping ticks, it is quite significant.

 

As was discussed in the earliest pages of this thread, this is all debateable and depends on the instrument in question.

 

For instance, most CME futures order books are FIFO, so there is no mechanism whereby HFTs can guarantee that they are ahead of you in the queue - they can't queue jump.

 

Secondly, there are restrictions on fill-to-cancellation ratios, and fines if a trader or HFT overshoots them (not an issue to pay the fines if you're trading massive size and making a lot of money - they're just another predictable fixed cost, I guess).

 

Then there is the point that as soon as market orders start to cross the spread hit the bid the HFT can tell whether the market will continue to decline (pulls it's orders) or whether the level will hold and a profit can be made . . . If it can do this (directional prediction) then why is the speed/limit order advantage even required?

 

Finally, supposing that the HFT "thinks" the level will hold as support and a profit can be had - it then decides to put a limit order into the orderbook somewhere overhead in order to take profit - it can only join the back of the queue when it does so, behind all the other orders already resting at that price.

 

I think the general gist of what you're saying is correct, but when it comes to the details and mechanics of what's going on, none of us really know . . .

 

BlueHorseshoe

Share this post


Link to post
Share on other sites
* * *

 

For instance, most CME futures order books are FIFO, so there is no mechanism whereby HFTs can guarantee that they are ahead of you in the queue - they can't queue jump.

 

* * * (Comment Edited)

 

BlueHorseshoe

 

This is an interetsing conversation. I haven't read all 9-10 pages of this thread yet, but had a few thoughts.

 

First, I think that scalping is alive and well and has just morphed into being done by HFT computers instead of locals in the pits. It used to be the locals could get better fills and were in-and-out with a few ticks before we could call-in to our broker, have him call the pit, place our orders and get confirmation numbers. Even if you did a lot of volume and had a dedicated phone line you were still slower that the locals. When markets went electronic, everyone effectively had the same shorter-time advantage and pit scalpers died out. Now electronic traders have to give way to HFT computer scalpers . . . who give way to those with faster computers . . . who give way to those with faster computers closer to the Exchange, etc.

 

Second, "scalping" is a relative term. Going for 1-2 ticks is scalping but so too is going for 5-10 ticks in a fast market like CL (Crude Oil) or GC (Gold). All I've done is raise my level of scalping above the noise generated by the HFT computers' whip in those markets. I can't compete with the HFT computers for 1-2 ticks, or even 5-10, but can take out 10-15 -- which I still believe is scalping.

 

But I disagree that the CME is still FIFO. At least without question. That's what was said for a long time before the existence of dark pools became known -- whereby HFT computers were allowed (for a price the Exchange didn't want to talk about) to "jump" the queue and get filled ahead of a limit order that had been sitting there a lot longer waiting for its fill.

 

I've traded CL for years and have a system to pick turning points in advance (sometimes). I've had my limit orders sitting there and can see the DOM as it comes into them and there is a total of, say, 5 at my level. Part of that is me. Price has way too often come into that level, hit it exactly and turned and I didn't get filled. I know I should have been #1 or #2 in the queue and can only guess I was "jumped" by a HFT computer. I can't prove it though.

 

Good Luck!

Chartsky

Edited by tradingwizzard

Share this post


Link to post
Share on other sites
This is an interetsing conversation. I haven't read all 9-10 pages of this thread yet, but had a few thoughts.

 

First, I think that scalping is alive and well and has just morphed into being done by HFT computers instead of locals in the pits. It used to be the locals could get better fills and were in-and-out with a few ticks before we could call-in to our broker, have him call the pit, place our orders and get confirmation numbers. Even if you did a lot of volume and had a dedicated phone line you were still slower that the locals. When markets went electronic, everyone effectively had the same shorter-time advantage and pit scalpers died out. Now electronic traders have to give way to HFT computer scalpers . . . who give way to those with faster computers . . . who give way to those with faster computers closer to the Exchange, etc.

 

Second, "scalping" is a relative term. Going for 1-2 ticks is scalping but so too is going for 5-10 ticks in a fast market like CL (Crude Oil) or GC (Gold). All I've done is raise my level of scalping above the noise generated by the HFT computers' whip in those markets. I can't compete with the HFT computers for 1-2 ticks, or even 5-10, but can take out 10-15 -- which I still believe is scalping.

 

But I disagree that the CME is still FIFO. At least without question. That's what was said for a long time before the existence of dark pools became known -- whereby HFT computers were allowed (for a price the Exchange didn't want to talk about) to "jump" the queue and get filled ahead of a limit order that had been sitting there a lot longer waiting for its fill.

 

I've traded CL for years and have a system to pick turning points in advance (sometimes). I've had my limit orders sitting there and can see the DOM as it comes into them and there is a total of, say, 5 at my level. Part of that is me. Price has way too often come into that level, hit it exactly and turned and I didn't get filled. I know I should have been #1 or #2 in the queue and can only guess I was "jumped" by a HFT computer. I can't prove it though.

 

Good Luck!

Chartsky

 

NYMEX / CME certainly claim that the CL contract is FIFO:

 

http://www.cmegroup.com/confluence/download/attachments/48627815/Globex%20Product%20Reference.xls?version=76&modificationDate=1381948777000&api=v2

 

BlueHorseshoe

Share this post


Link to post
Share on other sites
That kind of assumes that a scalper always has to be right no? Scalping is about talking advantage of a change in order flow, meaning they are just hopping in front of others to make a quick buck. No need to be so huge to reverse the market at all.

 

Scalping to me isn't "hopping in front of others to make a quick buck" to me it assessing direction and velocity and jumping on the train (possibly after others), but while the train is still going fast enough to get a few ticks before getting off. Volume isn't really as relevant unless you are huge.

Share this post


Link to post
Share on other sites

Join the conversation

You can post now and register later. If you have an account, sign in now to post with your account.
Note: Your post will require moderator approval before it will be visible.

Guest
Reply to this topic...

×   Pasted as rich text.   Paste as plain text instead

  Only 75 emoji are allowed.

×   Your link has been automatically embedded.   Display as a link instead

×   Your previous content has been restored.   Clear editor

×   You cannot paste images directly. Upload or insert images from URL.


  • Similar Content

    • By inthemoneystocks
      One of the most important reasons why traders take big losses is because they often fail to recognize when a trade has gone wrong. You see, stopping out of a trade is probably the biggest fault of traders and investors. Often, this happens to young and inexperienced traders and investors, but I know many veteran traders and investors that struggle with this as well. Early in my own career I struggled with stopping out of a bad trade myself, so I can sympathize with this problem. 

      The problem with taking a loss is really two fold. First, the trader has to admit that he is wrong. As you all know, as human beings we all hate to be wrong. The ego simply gets in the way and we all want to always be right all the time. The first secret in this business is to check the ego at the door. The market does not care about your the color of your skin, religion or anything else. It will move in the direction of the money and that is the bottom line. Once a trader or investor goes into what I call 'hope mode' the trade is over. I'm sure everyone has been in this position at one time or another. Simply put there is no room for ego or hope in the stock market. The market is always right and there is no reason to fight it. 

      Here is the second problem with taking a loss, it hurts. Pain and pleasure are the two reasons why humans do anything at all. As a human being, we are always looking to have pleasure and avoid pain. Well, losing money is painful and many people would rather simply hold a losing equity than lock in a small loss and move on. I cannot tell you how often I see a trader hold a losing trade only to see the position move further out of the money. Many years ago I watched a day trader blow up a $200,000 account in a single day averaging in on a bad day trade. To this day I can remember the look on his face as his money vanished in thin air. Believe it or not, this trader could have exited the position with a $500.00 loss, but instead he kept averaging in and fighting the position until he was wiped out. As a rule, once you have your full position you should never average in on a trade. At that point, it is critical to know where your max loss is going to be and stop out if that level is breached.

      Now when should we stop out? The answer to this question is not that simple, but here is what I personally do. I always place my stop loss below an important breakout or pivot on the chart. You see, prior breakout or pivot levels are usually defended when retested. After all, this is usually an area where institutional traders and investors got involved, that is why there is a pivot low or high on the chart to begin with. If that level is breached on a closing basis then I will move out of the position. So If I took a trade based on a daily chart pattern then I will usually check the daily and weekly chart levels. If there is a major pivot on the weekly chart then I will use a week chart close as my stop out level. While this method may not be perfect, it has saved me from much bigger losses when I have been wrong.



        Nicholas Santiago
    • By trading4life
      Hello, My name is trading4life.
      I just joined this forum.
  • Topics

  • Posts

×
×
  • Create New...

Important Information

By using this site, you agree to our Terms of Use.