Daytraders - Do You Know Your Enemy?

[Predictor 11]

One of the problems with the "go it alone" approach is that without a structured program, one can spend a long time running in circles. I believe the ops method actually makes a lot of sense. Of course, it is not the way to develop the type of "read" I've developed. But a structured program, i.e systematic study may reveal unique insights that others don't understand or know.

 

One caveat is to understand that most statistical measures don't faithfully apply to the markets. They can still prove useful but without a strong statistical background then it may be difficult to understand everything. For example, correlation only models linear versus non linear relationships. Many functions assume the normal distribution. Many functions assume a stationary time series. Beyond the usual reservations, I find that even basic statistical measures can sometimes prove useful. Also, if there are certain laws that govern these measures that aren't understood then it is even more difficult, i.e without a framework. A degree in randomness be would useful.. I imagine.

 

I would say regardless of whether you go it alone, try to build a system, or intend to "stare at quotes" that one tries to build some sort of systematic and structured program.

[BlueHorseshoe 10]

A common misconception - probably initiated and perpetuated by the HFT crowd which attributes the reduced spreads to HFT rather than the real cause - 'decimalization'.

 

It was decimilization that was responsible for the decreased spreads in Stocks by 2002 when HFT still only acccounted for about 10% of the market.

 

http://www.gao.gov/new.items/d05535.pdf

 

FWIW, HFT doesn't increase liquidity either. It increases volume - not the same thing - especially in melt-ups/downs.

 

I don't really know anything about stocks, but HFTs are clearly a presence in the ES (for whatever purpose) - does anyone ever remember this contract consistently trading with anything more than a 1 tick spread during normal volatility?

 

BlueHorseshoe

[Predictor 11]

A clue might be the volume and liquidity drops off significantly after hours. .. whatever they are doing they want the cash market open. There is a cancellation ratio/fee in futures that doesn't apply to stocks...

 

I'd imagine that many HFT's want to capture the spread.. I imagine the very best orders would never move the market, right.. Small orders... Any LQ provider needs to manage inventory, needs to manage directional exposure, and needs to be able to maintain good queue position in the book...

[Paul-TC 10]

I don't really know anything about stocks, but HFTs are clearly a presence in the ES (for whatever purpose) - does anyone ever remember this contract consistently trading with anything more than a 1 tick spread during normal volatility? BlueHorseshoe

 

The spread has never been an issue on the ES. I began trading it in 1999 on a dedicated Globex Terminal and it's always had a 1 tick spread during normal volatility. That was true even when the order size was limited to a max of 30 contracts!

[BlueHorseshoe 10]

The spread has never been an issue on the ES. I began trading it in 1999 on a dedicated Globex Terminal and it's always had a 1 tick spread during normal volatility. That was true even when the order size was limited to a max of 30 contracts!

 

Thanks.

 

So in other words, HFT can take co credit for narrowing the spread of the ES because it has always been as narrow as it can be . . . 30 contracts is nothing to an institution nowadays - was the instrument originally created more with smaller traders in mind?

 

I know plenty about the history of commodity futures, but almost nothing about the origins of the e-minis.

 

BlueHorseshoe

[BlueHorseshoe 10]

There is a cancellation ratio/fee in futures that doesn't apply to stocks...

 

Please can you explain this?

 

BlueHorseshoe

[Predictor 11]

There is a cancellation ratio that is set by the CME. It is ratio of cancelled vs placed trades. If a trader cancels too many trades then they can be fined. Many HFT programs are suspected of placing and cancelling a high number of trades.

 

As noted there are non primary liquidity providers (traders), as well.. This means that anyone trying to profit from the spread will need to be able to keep enough orders out to ensure good queue position... moving the order will push one into the back of the queue.

 

 

 

Please can you explain this?

 

BlueHorseshoe

[MightyMouse 12]

One of the problems with the "go it alone" approach is that without a structured program, one can spend a long time running in circles. I believe the ops method actually makes a lot of sense. Of course, it is not the way to develop the type of "read" I've developed. But a structured program, i.e systematic study may reveal unique insights that others don't understand or know.

 

One caveat is to understand that most statistical measures don't faithfully apply to the markets. They can still prove useful but without a strong statistical background then it may be difficult to understand everything. For example, correlation only models linear versus non linear relationships. Many functions assume the normal distribution. Many functions assume a stationary time series. Beyond the usual reservations, I find that even basic statistical measures can sometimes prove useful. Also, if there are certain laws that govern these measures that aren't understood then it is even more difficult, i.e without a framework. A degree in randomness be would useful.. I imagine.

 

I would say regardless of whether you go it alone, try to build a system, or intend to "stare at quotes" that one tries to build some sort of systematic and structured program.

 

Your quotes ("") indicate that you take issue with the possibility of being able to understand orderflow through tape reading in a short period of time. Sour grapes aside, some things are easily overlooked when they are over-analyzed. If you came across something similar or better after your two year period, then all that matters is that you did find it and that is great.

[SIUYA 11]

A common misconception - probably initiated and perpetuated by the HFT crowd which attributes the reduced spreads to HFT rather than the real cause - 'decimalization'.

 

It was decimilization that was responsible for the decreased spreads in Stocks by 2002 when HFT still only acccounted for about 10% of the market.

 

http://www.gao.gov/new.items/d05535.pdf

 

FWIW, HFT doesn't increase liquidity either. It increases volume - not the same thing - especially in melt-ups/downs.

 

yep - but who helped push decimalization, and then by their nature really pushed the spreads in, and helped open access to the markets for most people - chicken and egg really, but even with the regulations/incentives in place you still need people to push it. HFT in a simple form is just a super fast way of doing what humans would do naturally. The HFT before computers where the pit scalpers and market makers - nothing here has changed. Its a bit like saying that the first computer, hell the first abacus was really the cause.......It was always going to be a natural progression - diming and speed - once you are allowed to. (by the way Australia has had decimalization and very close spreads in liquid stocks well before the floors closed)

 

....not sure how volumes have increased but liquidity has not.......unless of course volatility has massively increased, in which case vwaps and other such measures are designed to minimise this.

Mind you volumes have increased in the previously high volume instruments - again by their very nature this makes sense, as they are looking for volume and liquidity.

One issue the exchanges have is how do they increase the liquidity and decrease the spreads in the less traded instruments - personally I dont think the exchanges care - and one of the big issues is the mismatch of incentives driving the exchanges and their buddying up with the HFT. ---- previously they could enforce markets, but the spreads were wide and when the s..t hits the fan every one ducks for cover.....its the nature of markets.

 

plus when it comes to meltups/downs - they may exacerbate them some what but it is no news flash that these occurred with zero liquidity before HFTs, before computers - they are often called gaps, crashes and manias.

 

I think people are so worried about the HFT clipping them they are missing the advantages they have brought (along with regulatory changes) and more importantly they have missed the problems that may occur as a result......now the issues involved with it being an uneven playing field with hidden orders and different order types - now that is criminal.

Edited August 10, 2012 by SIUYA

[Paul-TC 10]

Thanks.

 

So in other words, HFT can take co credit for narrowing the spread of the ES because it has always been as narrow as it can be . . . 30 contracts is nothing to an institution nowadays - was the instrument originally created more with smaller traders in mind?

 

BlueHorseshoe

 

I presume you meant 'no credit' rather than 'co credit'. If so, I agree and inappropriately giving HFT the credit for narrowing the spread applies to Stocks and Options too.

IMO, HFT supporters perpetuate this myth to justify its existence.

 

I don't believe the order limit was in place because the eminis were created for the smaller trader. It was always designed to succeed the pit-traded SP eventually.

 

The limit (which was gradually raised to 100, then 250 and now almost unlimited) was put in place while the CME ironed out a few teething problems on Globex. The limit prevented 'sweeping the book' by the bigger players until the engine was stable and the volume traded grew. Most were still trading the SP back then and phoning their orders in. Many of them were resolute in refusing to 'go electronic' so the volume took a while to pick up but the spread was still 1 tick under 'normal' conditions. Worth pointing out that I never experienced more than 2 tick slippage (although I accept that there must have been times that I might have been), during the 'flash crash', despite HFT, the spread grew to 10 full handles (40 ticks!).

 

I should clarify that HFT, if taken to be no more than its name implies, could never be a problem. How fast and frequently 'legal' trades are executed is irrelevant.

 

It's the fact that they are being allowed to enage in tactics that are illegal according to Securities Law. e.g. Orders are often being placed with the sole intent of NOT being executed and are rendered not executable due to the speed with which they are pulled. Layering, spoofing, quote stuffing, wash trades - all of which are illegal - occur every day but only in rare cases are they punished with paltry fines and permitted to do so with no admission of any guilt. e.g. High-frequency trades earn $2.3m fine - FT.com

 

It is also troubling that these volumes seem to be growing with no increase in margin being placed to cover the positions. e.g. Knight Capital Group acquired $7 Billion in Stocks at one point when it had less than 10% of that in cash for margin?

 

If there is payment to HFTs for trades, it will encourage wash trades, quote stuffing, spoofing et al. If you add to that no limit on how much can be traded (i.e. no margin requirements), strangling the market in 4 tick ranges (sometimes for hours) while 100's of thousands of contracts 'appear' to be legitimately traded is going to be lucrative. And that's exactly what we see every day soon after the open. It started in about 2003 during which I recorded the lowest RTH range of 3.25 points in the ES.

 

As with everything 'bank' related these days, it isn't the lack of regulation that's the issue, it's the woeful lack of enforcement.

 

Of course, it is still possible to make money in the ES (even easier in other markets) but price 'discovery' has been replaced with price 'control' and such strangulation of the market in such narrow ranges for extended periods removes a lot of trading opportunities that used to exist.

 

By way of coincidence, Denninger just posted an article on his website yesterday about this same HFT issue.

 

HFT Mess -- A *SIMPLE* Answer in [Market-Ticker]

[Paul-TC 10]

yep - but who helped push decimalization, and then by their nature really pushed the spreads in, and helped open access to the markets for most people .

 

Nonsense, it doesn't matter who pushed for decimilization. It was the decimilization that produced the reduced spreads - not the HFT process itself - as you originally asserted - which only started to dominate waaay after the spreads had already narrowed.

 

If there was another group (let's call them the LFT (Low Frequency Trading - or even the cake-baking-grannies of Minnesota) who pushed for decimalization - and achieved it, the result in spreads would have been the same but would have had nothing to do with how infrequently they trade or how they make their cakes.

[Paul-TC 10]

Sorry, I should have added this to the previous post but I hit enter too early and I'm unable to edit because I'm new here and my posts are still being monitored before being added.

 

HFT in a simple form is just a super fast way of doing what humans would do naturally. The HFT before computers where the pit scalpers and market makers - nothing here has changed. Its a bit like saying that the first computer, hell the first abacus was really the cause.......It was always going to be a natural progression - diming and speed - once you are allowed to.

 

You are describing the advantages gained with the advent of electronic trading over the old, manual method - and you are absolutely correct. It was a truly wonderful thing. Electronic markets did originally level the playing field as you point out. Unfortunately, you then spoil it by mistakenly attributing that improvement to HFT - with zero consideration of the very real fact that access, speed of execution and spreads had nothing to do with HFT and that they all improved before HFT's dominance.

 

In short, you are continually, comparing post-HFT with pre-Electronic/pre-Decimalization when the only relevant comparison, for the pruposes of this discussion, is post-HFT(dominance) with post-Electronic but pre-HFT(dominance)

 

(by the way Australia has had decimalization and very close spreads in liquid stocks well before the floors closed).

 

...... and also well before HFT - which is my point.

 

....not sure how volumes have increased but liquidity has not.......

 

Because the volume is mostly an illusion, fleetingly displayed (if at all) but unexecutable, or 'washed' by buying/selling their own orders, ad nauseum, for the liquidity rebates given by the Exchanges. Google is your friend and "Whats the difference between Volume and Liquidity HFT" will be a much-needed education for you - and just in the first 4 articles.

 

Similarly, unquestioned belief in the myth put forth by HFT about their responsibility for the narrowing of spreads is not supported by studies done by Nanex : Nanex ~ The Tighter Spreads Lie

 

.....and, according to this study, may even be detrimental to those spreads that so many mistakenly believe have narrowed because of HFT: High-Frequency Trading Is Costing Investors Billions: Pragma Securities - Advanced Trading

 

As I mentioned in another post, how quickly legal trades are executed is not the problem. If that's all HFTs were doing, we wouldn't have a problem - but then they probably wouldn't exist to any great extent as their risk would be too great without the rebates, permitted spoofing/layering, etc. etc.

 

However, it is utter garbage to attribute access, 'fairness' and spread gains to HFT when it should be obvious, and is demonstrably provable, that it's the result of decimalization and electronic trading.

 

....plus when it comes to meltups/downs - they may exacerbate them some what but it is no news flash that these occurred with zero liquidity before HFTs, before computers - they are often called gaps, crashes and manias........

 

No one ever suggested, or even thinly implied, that they didn't occur before HFTs or computers. I don't see why you feel the need to address the suggestion as if it had been made.

 

But as you brought it up... : It's comparatively early days in HFT's history and so we may have too small a sample size to make any definitive conclusions. However, what can be observed so far is that their occurrence has been more frequent, more violent, with wider spreads and lower liquidity than ever seen before.

 

Admittedly, HFT never promised to eliminate these events but its record so far is the complete opposite of everything that HFT claimed would dramatically improve.

[BlueHorseshoe 10]

I think people are so worried about the HFT clipping them they are missing the advantages they have brought (along with regulatory changes) and more importantly they have missed the problems that may occur as a result......now the issues involved with it being an uneven playing field with hidden orders and different order types - now that is criminal.

 

What I hoped would emerge from the thread is not so much a fear of specific effects of HFTs (the thread isn't HFT specific, by the way!), as much as a healthy respect for the capabilities of other market participants operating in this timeframe.

 

Maybe this is just a symptom of a lack of sense of self-worth in my own capabilities as a daytrader?

 

Many buy-and-hold investors are very well aware that their research capabilities are dwarfed to many magnitudes by those of, say, Berkshire Hathaway. Reading the forum I suspect that few daytraders ever pause to consider that they might be seriously 'out-gunned' by other intraday participants. There will always be someone much, much better (capitalised/networked/tech'd-to-the-eyeballs) - this doesn't necessarily leave you at the bottom of the food chain, but surely it makes sense to have some understanding of where you are in the food chain?

 

 

BlueHorseshoe

[BlueHorseshoe 10]

Staying with the ES . . .

 

When an iceberg order reloads on Globex, it goes to the back of the queue.

 

The value of this information is that, though the size of the iceberg is unknown (unless you deploy an algorithm capable of re-generating it), an iceberg has no special impact on judging one's own position in the order queue.

 

BlueHorseshoe

Edited August 11, 2012 by BlueHorseshoe

[Predictor 11]

It is possible to estimate your own position in queue, obviously. There isn't 100% way to know however because orders underneath your own can be pulled while simultaneously new orders can be added at the back.

 

Fx.. you are at 1000, 500 pulls, and 500 adds... the order book can many times per second...

 

Icebergs do not reload (typically) on a loaded order book. They reload at the end. That's the purpose is to disguise how many orders are for offer/bid at a level. A trader might decide to reload 50 whenever the offer drops below 50 lots. When the bid/offer becomes imbalanced then it encourages traders to go to market in front of the imbalance. There is risk to trying that though because the bid may drop before the market is executed. Likewise, you can't just grab the price if it ticks down because your limit is going to be at the back of the queue...

 

For a myriad of reasons.. I think it will be difficult for retail trader to operate in this space. I think if you are interested in HFT that you are better off looking at stocks, rebate strategies, NASDAQ (I hear they give HFT's special information on WHO the order was from), and possibly Forex where semi-martingale strategies could be tried.

 

Staying with the ES . . .

 

When an iceberg order reloads on Globex, it goes to the back of the queue.

 

The value of this information is that, though the size of the iceberg is unknown (unless you deploy an algorithm capable of re-generating it), an iceberg has no special impact on judging one's own position in the order queue.

 

BlueHorseshoe

Edited August 11, 2012 by Predictor

[BlueHorseshoe 10]

It is possible to estimate your own position in queue, obviously. There isn't 100% way to know however because orders underneath your own can be pulled while simultaneously new orders can be added at the back.

 

Now then, Mr Predictor . . .

 

You're absolutely correct and, true to you alias, you're a few steps ahead of where I'm trying to get to here.

 

If you can give me a few days to catch up with you (in detail and with examples) for anyone now or in the future who reads this thread and doesn't already know what you're talking about, then that would be greatly appreciated.

 

You're tagged as a vendor, pressumably of educational material, so I'm sure you'll understand my desire to progress through this in a measured manner - the purpose of this thread was to share information as I learn it myself . . .

 

Regards,

 

BlueHorseshoe

 

ps - thanks for the posts on cancellation ratios - sounds incredibly stupid but never having daytraded futures I was completely ignorant of them - reading up now!

[Blaiserboy 10]

BlueHorshoe, Predictor

 

 

There is a lot of information being passed on in this thread..... I am learning quite a bit.

 

Please post more and more, I am sure others are benefiting as well

 

Thanks

[Paul-TC 10]

Staying with the ES . . .

 

When an iceberg order reloads on Globex, it goes to the back of the queue.

 

The value of this information is that, though the size of the iceberg is unknown (unless you deploy an algorithm capable of re-generating it), an iceberg has no special impact on judging one's own position in the order queue.

 

BlueHorseshoe

 

Iceberg orders can differ according to the Exchange. Some Exchanges don't hold them as native orders at all and, if required, they have to be managed locally instead. Even Icebergs that are held natively, don't reload under the same rules. So check carefully with each Exchange directly if Iceberg orders are important to you.

 

As you're specifically asking about the ES on Globex, which operates a FIFO hierarchy order system, each 'MaxShow' does go to the back of the queue when it's reloaded - as you said.

 

So, I think what you're asking is: If there's an Iceberg order for 500 with only 10 showing at any time and you add, say 20, to the queue while only 10 are showing on the book - (which will now show 30), are you in positions 11-30 or positions 501-520?

 

You are in positions 11-30.

 

If 10 are traded, 30 will still show (after the Iceberg order reloads another 10), but as the new tranche reloaded from the Iceberg order has a new, later timestamp, it is subordinate to your order and so you will now be in positions 1-20.

 

Note: Native Globex Iceberg orders are not refreshed according to Bid/Offer levels reaching certain levels on the book. They are only refreshed after each 'MaxShow' tranche has been traded in full. To reload according to any other criteria can only be done programmatically at the local level rather than by the Exchange.

[BlueHorseshoe 10]

Iceberg orders can differ according to the Exchange.

 

Thanks for the confirmation Paul.

 

I've been focussing more on learning about how orders are handled by the CME, as this is where my own interests lie. If you're able to fill out the picture with a little of how other exchanges deal with icebergs then that would be greatly appreciated.

 

Cheers!

 

BlueHorseshoe

[SIUYA 11]

Hi Paul TC, I think we agree on a lot of things here, and decimalization is definitely a factor, as were the NBBO rules the rise of the computers and access to markets among many of the things. While you may be right that HFT (in its current form) has not decreased spreads further, and is now causing a lot of problems, HFT participants in the past were definitely a factor/participant in reducing spreads and costs.

 

This is not to get into an argument for arguments sake, I would also like to better understand HFT as it currently is - (and maybe Blue this is not the thread for it (if you think that then a new thread might be one to discuss the workings/merits/history of HFT)) - but there are some things that for me, maybe be simple definition differences, or versions of history..... so maybe you could give more light on a few things. Or if we clarify a few things as you see them.....(I don’t care either way for HFT, but I do think the complaints against HFT are just as misdirected)

 

HFT trading has always been around in some form and for the other reasons of quoting in 1/16ths, lack of access to markets, the protection of floor traders and spreads by the exchanges it was the new participants once given the ‘tools’ that naturally reduced the spreads. The old guard was trying to protect their patch. It was not the mums and dads, nor generally the institutions that buy and sell throughout the day who keep spreads in. It is the HFT….and yes they can have co credit, and clearly not all the credit (and maybe I phrased myself poorly in that way – but stick with me here). You may have all the rules and regulations, but if you don’t have the willing and profitable participants then nothing would happen. So while HFT are not the only reason I don’t think people can say they have no value, and while they complain about the ‘benefits’ of HFT not being benefits and that they contribute nothing, its hard to take any discussion on things seriously.

Wall st /the exchanges was dominate and now that HFT is dominate people are complaining about the same issues – I was clipped, my spread is too big, these guys are front running me…..same issues different participants....when the real issues might be different.

 

Now…..clearly the HFT has morphed into something different…..as the Nanex guys show its only from 2007….and so I guess you could say these new ultra HFT traders are causing real issues, and ultimately I think this is where a lot of the distinction needs to be made between just HFT, and these HFT now. I think you nailed it with the quote – “when the only relevant comparison, for the purposes of this discussion, is post-HFT(dominance) with post-Electronic but pre-HFT(dominance)”

 

Re the Volume illusion point….quotes which the Nanex guys talk about and the actual traded volume as a percentage of the quotes is the issue. Volumes have increased, and while they may have decreased since the post 2007 highs and GFC periods, they are still higher – a result of the combination of HFT, computers, decimalization etc; Quotes have always been illusionary. Trading volumes are where you wish to trade and actually trade are important, and limit orders still get hit. (The two tiered system is another issue). I still have it yet explained to me how this is any different apart from speed, and volumes of quotes as to what has always been done. Its just front running trying to pick up where the orders are. (as opposed to market making, arbitraging etc)

 

So if you wish to say that it is utter nonsense that the HFT traders don’t contribute as participants then I think you are throwing everything out and blaming other market participants for not adapting. It is equally nonsense to say they don’t contribute. Nobody seemed to mind when the HFT traders did and do provide liquidity, The Pragma securities article is a bit of sour grapes – they are basically saying ‘we were once liquidity providers but now we have to cross the spread to get set because other people are doing the trades in front of us……this is costing us.’ Even they say changing things would lead to reductions in volumes…..so it is definitely an issue of them protecting their own interests first. (no different to the HFT). The dominance of HFT as front running is of course concerning.

 

I 100% agree with you with the issues regards extra order types, legal trades, spoofing and the maker-taker model s, and I also think that people are missing the real issue that may occur as a result of the HFT. Instead they are blaming them for front running their orders, and this is a side show in which the participants may have changed, and new issues have been brought up – the correlation of markets, the speed of contagions and errors, and the effects and looping of these errors….as Knight, and the flash crash has shown.

 

Current HFT may have other issues that we can agree on, and the debate may get sidetracked from those – especially at the national levels, and getting all hot under the collar about claims from one side or the other, does not help.

 

I thought this was the best article of them…

http://www.sps.ed.ac.uk/staff/sociology/mackenzie_donald/?a=78186

[BlueHorseshoe 10]

This is not to get into an argument for arguments sake, I would also like to better understand HFT as it currently is - (and maybe Blue this is not the thread for it (if you think that then a new thread might be one to discuss the workings/merits/history of HFT))

 

Hi SIUYA,

 

I have no objections whatsoever to this being discussed here - it all seems relevant to me.

 

It's great that the thread is attracting contributions from such a varied range of people, and that people are willing to share their knowledge without trying to belittle anyone else with it.

 

The only objection that I do have is all these links to pdfs - my printer has run out of ink and my eyes are suffering from screen-burn!

 

Thanks,

 

BlueHorseshoe

[Paul-TC 10]

While you may be right that HFT (in its current form) has not decreased spreads further, and is now causing a lot of problems, HFT participants in the past were definitely a factor/participant in reducing spreads and costs.

 

While it's very easy to state that HFT was 'definitely a factor/participant in reducing spreads and costs' - and then present it as an incontrovertible truth, I've never seen any hard evidence to support it.

 

However, I have seen results of studies that show that they haven't been a factor. In fact, some of them (like one I posted a link to) go further and conclude that HFT has actually increased the costs. I accept that you may not need the same level of supporting evidence to reach your 'definite' conclusions about HFT's claims as I do, and you have every right to, but not everyone 'does' unquestioned faith when faced with unsubstantiated claims.

 

I also accept that The Pragma Securities article 'might' be sour grapes. I really don't know that it is - but as you don't either, I'll make that issue easy for both of us and assume it is. However, it's irrelevant. Sorry, but ad hominem attacks of messengers while conspicuously avoiding the results of their discoveries never sits easy with me.

 

e.g. If Ted Bundy claimed that his investigations revealed that Jeffrey Dahmer was a murdering psycho, does it make Jeff any less guilty because of Ted's 'sour grapes' ?

 

All that matters is: Were the findings of their study accurate? I'm more than happy to listen to anyone who has a counter-study that debunks their findings but I haven't seen one. All I'm hearing in response is equally unsubstantiated speculation about 'sour-grapes' as the reason for their study and that is not an acceptable substitute for an opposing study.

 

Re the Volume illusion point….quotes which the Nanex guys talk about and the actual traded volume as a percentage of the quotes is the issue. Volumes have increased,

 

No one said volume hasn't increased - but again, volume is NOT the same as liquidity - and that is what has dried up. Actual traded volume as a percentage of the quotes is not the only issue either.

e.g. If you and I trade 1000 shares of a $10 share back and forth 100 times (or I wash those trades myself 100 times), the volume on the tape 'appears' to be 100K but, at best, the real volume is only 1K. The trades provide no liquidity because the HFT machines ensure that they can never be traded by anyone else - but I/we would get rebates from the Exchange.

 

You may have all the rules and regulations, but if you don’t have the willing and profitable participants then nothing would happen. So while HFT are not the only reason I don’t think people can say they have no value, and while they complain about the ‘benefits’ of HFT not being benefits and that they contribute nothing, its hard to take any discussion on things seriously.

 

I don't really know what to say. You've said nothing in that paragraph but have just complained about the complainers? How does that seek to discover the truth?

Using the same principle, one could just as easily reconstruct that paragraph and say:

 

"One could have all the fraud and crime but if you don't have willing and honest participants seeking to expose it then nothing would happen. So while HFT 'might' be blatantly defrauding investors, I don't think the unwilling people can say that HFT has 'definite' value when they provide only purely subjective opinions about that value but then complain about any entity that endeavours to provide more stringent tests as being 'sour grapes' - without providing even one iota of evidence that their studies are wrong".

 

Yes, it is very hard to take any such discussion seriously.

 

Even the article you posted a link to states:

 

"Spreads in the U.S. have come down rapidly in recent years, often falling to the lowest possible value in the current pricing regime, one cent. Although multiple factors are

involved – most obviously the reduction in 2001 in minimum increment from one-sixteenth of a dollar to one cent – the growth of electronic market-making has almost certainly led to lower spreads"

 

So, it attributes the reduction to decimalization in 2001 (before HFT dominance) and electronic trading. It makes no mention of any studies showing that HFT 'caused' that lower spread as you originally asserted. Furthermore, like you, it just 'states' that HFT did without providing one shred of evidence.

 

I have said this twice already but I'll say it again. There is nothing wrong with just legally executing trades with a high frequency - or even a very high frequency - but that is not what they're doing.

 

They are not legally executing trades and the high frequency with which they are able to pull, spoof, quote stuff etc. does not benefit anyone but them. There is no evidence, at least I still haven't seen any, that HFT has - or even could - reduce spreads beyond what decimalization and electronic trading in general has already done.

 

By not being able to show (with hard evidence) that there is a benefit to the lay investor/trader, one must, at the very least, consider that there may not be one.

 

If in fact, there is no benefit and if it exists purely to facilitate a crime (There's no 'if' anymore really, It has already been proven that they are breaking several Security Laws on a daily basis - all that is in question now is whether it is 'purely' for that purpose), then unquestioningly defending its wholly unsubstantiated claims of spread (or any other benefit) only further empowers them and enables the crime.

 

I don't believe in just attacking HFT for the sake of it but if repeatedly being caught breaking the law doesn't cause you to question their (and your own) unproven claims of their benefit, then what will?

Edited August 12, 2012 by Paul-TC

[Paul-TC 10]

Thanks for the confirmation Paul.

 

I've been focussing more on learning about how orders are handled by the CME, as this is where my own interests lie. If you're able to fill out the picture with a little of how other exchanges deal with icebergs then that would be greatly appreciated.

 

Cheers!

 

BlueHorseshoe

 

Wow, I wish I was an expert on every order type of every Exchange:rofl:

 

I think you'd have to check with each Exchange - and even each Instrument may be treated differently but one of the main differences is that not all exchanges operate a FIFO hierarchy. The order in which trades are filled is according to a ratio algorithm - Eurex being the most well-known example I guess.

[SIUYA 11]

While it's very easy to state that HFT was 'definitely a factor/participant in reducing spreads and costs' - and then present it as an incontrovertible truth, I've never seen any hard evidence to support it.

 

As I said before Paul - you are taking HFT as it currently exists....post 2007.

Now if you have watched people who trade a lot of volume in this manner, and would be generally considered HFTs they continually need to be in front of the que.....it is part of the strategy.

 

HFT by its very nature made the most of the tools available to it and the fact is they now dominate the market and the spreads are very tight. (But this seems to be what you are debating) I am not sure that some academic paper is going to convince you if the market results dont. Now yes HFT may be devouring itself, and maybe doing illegal trades as you keep insisting - best talk to the SEC, as if you like evidence, then surely they would be stopping it.

Plus these are not my definitive conclusions - you may have reached some, I am trying to better understand what happens now in HFT without the PR and sour grapes.

 

Re the costs - again are you talking about HFT as they currently exist? You seem to refer to studies that come from 2007 onwards for your definitive claims. So while I accept there is a lot of sour grapes, and a lot of PR from both sides, I am not defending them as they exist but surely its not just a co-incidence that as HFT in pre and post 2007 has come to dominate the markets and do the volume, and costs have come down (check an old brokerage statement from 15 years ago).....is is this ALL because of computers and decimalization.

 

 

I also accept that The Pragma Securities article 'might' be sour grapes. I really don't know that it is - but as you don't either, I'll make that issue easy for both of us and assume it is. However, it's irrelevant. Sorry, but ad hominem attacks of messengers while conspicuously avoiding the results of their discoveries never sits easy with me.

 

the results of their discoveries are what - that it costs them more money as they now have to cross the spread when previously they did not have to.

Wow - what an insight - a bit like the horse owners complaining about those dang noisy cars. (much like your Ted Bundy comparison - this is not about guilt - unless of course such studies are saying the HFT traders are stealing our (whoops I mean your) money - whereas previously we used to do it )- so no one is disputing their finding that it may be costing them more. A bit like Walmart underselling people.......

I am not attacking the messenger, but its a bit tough to get criticism from some one for saying you believe the HFT PR, when here I will prove with my own PR you are wrong.

 

My point is that the focus should be on the real issues - not 'its costing me more' - but those of the risks of the current HFT. You seem to want to demonise the HFT with your own PR

 

No one said volume hasn't increased - but again, volume is NOT the same as liquidity - and that is what has dried up. Actual traded volume as a percentage of the quotes is not the only issue either.

e.g. If you and I trade 1000 shares of a $10 share back and forth 100 times (or I wash those trades myself 100 times), the volume on the tape 'appears' to be 100K but, at best, the real volume is only 1K. The trades provide no liquidity because the HFT machines ensure that they can never be traded by anyone else - but I/we would get rebates from the Exchange.

 

Re the liquidity v volume and quoting.

Maybe I am being a little too simplistic in my assessments......but I have seen this over the years in many forms. Plus as an ex-equity options market maker that used to constantly deal with people making the same conclusions you have reached.....and if you apply it to many situations in real life.....quotes are largely an illusion.

Every had a real estate agent tell you what your house is worth. Ever made a market in something yourself. (In options you used to love those people who wanted to trade at a theoretical fair value price, and then complained if they did not get set) Ever tried to raise money and then go back to the person who said 'yeah i will seed your fund'. Are the HFT under any obligation to set other traders?

 

Now again I will repeat myself.....if the issues are that the HFT are doing something illegal then its a different matter, or the playing field is not level (and it appears its not with the HFT special orders), or they are stuffing quotes, front running and other such market manipulation practives then yes focus on that - not whinge and cry when 'that person renigged on a quote because they had a faster machine than mine', or I did not want to pay the offer, or sit in the machine on the bid, or participate when the sellers were smashing the stock even though I am a long term investor. Isn't this one of the reasons for why dark pools originated?

 

I don't really know what to say. You've said nothing in that paragraph but have just complained about the complainers? How does that seek to discover the truth?

Using the same principle, one could just as easily reconstruct that paragraph and say:

 

"One could have all the fraud and crime but if you don't have willing and honest participants seeking to expose it then nothing would happen. So while HFT 'might' be blatantly defrauding investors, I don't think the unwilling people can say that HFT has 'definite' value when they provide only purely subjective opinions about that value but then complain about any entity that endeavours to provide more stringent tests as being 'sour grapes' - without providing even one iota of evidence that their studies are wrong".

 

Yes, it is very hard to take any such discussion seriously.

 

The dreaded word of the truth - Sorry Paul, hard to criticise me when I never claim to state the truth, but you want to, and again, I have referred back to the point that you have taken the recent HFT dominance as the measure from which to measure and then pour negativity on HFT as if it never existed prior to 2007 and its current dominance.....and I am happy to agree with you.

 

Plus then we get to your version of the truth - the illegality of their trades......ok....as I have constantly said, I think this is an issue and if its proven that they are acting illegally you had best talk to the SEC. These are real issues......not people who complain that their business is being reduced due to people who are doing something they used to.....remembering that they too are probably HFT.

(Cynic as I may be it pretty easy to see when someone is pushing their own barrow, next we will have the French imposing transaction taxes to punish the speculators - the laws of unintended consequences)

 

I have said this twice already but I'll say it again. There is nothing wrong with just legally executing trades with a high frequency - or even a very high frequency - but that is not what they're doing.

 

They are not legally executing trades and the high frequency with which they are able to pull, spoof, quote stuff etc. does not benefit anyone but them. There is no evidence, at least I still haven't seen any, that HFT has - or even could - reduce spreads beyond what decimalization and electronic trading in general has already done.

 

By not being able to show (with hard evidence) that there is a benefit to the lay investor/trader, one must, at the very least, consider that there may not be one.

 

If in fact, there is no benefit and if it exists purely to facilitate a crime (There's no 'if' anymore really, It has already been proven that they are breaking several Security Laws on a daily basis - all that is in question now is whether it is 'purely' for that purpose), then unquestioningly defending its wholly unsubstantiated claims of spread (or any other benefit) only further empowers them and enables the crime.

 

I don't believe in just attacking HFT for the sake of it but if repeatedly being caught breaking the law doesn't cause you to question their (and your own) unproven claims of their benefit, then what will?

 

Actually why bother.......the banks have been continually breaking the laws for years, are constantly fined for doing so and settle without criminal charges and continue on.......

 

 

Paul - we are agreeing on a lot of things here - I am not sure if you can see past the PR on that, and if it makes it easier I will concede to the world - I was wrong HFT have never contributed to reducing costs or spreads....I will go one better if it makes it even easier - HFT are increasing costs, thieving bastards who take our money and they took our jobs, and corrupt the youth.....better now?

[BlueHorseshoe 10]

Take a look at the first image attached below.

 

This shows the results of a strategy traded on a 1 min chart of the ES using a single contract. I have chosen this timeframe because clearly it is not high frequency (and is therefore of relevance to 'retail' daytraders), but quickly generates a large number of trades over which to gauge performance. The test spans five consecutive days. No commissions were deducted, but with the system generating over 100 roundtrips per day, fees would have relatively little impact on the average profit per trade shown below.

 

Net Profit: $5,450

Profit Factor: 3.32

Percentage Winners: 71.03%

Maximum Drawdown: $275

Average Trade Profit: $10.19

 

The important thing that you need to know is how I have instructed the back-testing software to simulate fills. Limit orders have been used for both entries and exits, and the software has filled all orders at the limit price as soon as that price is touched, which is to say, as soon as a single trade occurs at that price. This is not a realistic representation of what would happen in normal trading. Not even remotely so. So, kids, don't try this at home!

 

Whether a limit order is filled depends on how close to the front of the queue (at that price level) it is. If every limit order in the queue at a particular price level is matched with a market order, and at least one order at the subsequent price level is also matched, then the market will move by one tick.

 

In the back-test, one simple way to get an idea of which of the limit orders would actually have been filled is to specify that the software simulate a fill only when the market trades through the limit price. The second image shows the equity curve once this specification is applied.

 

As you no doubt already know, a process of adverse selection is at work here; not all trades that would become winners are executed (and somewhat ironically, the more accurate the strategy is able to pick the market’s turning point, the fewer winning trades will be filled), whereas all losing trades will be filled, as a loser must, by definition, trade through the limit price.

 

This second approach isn’t entirely helpful either, providing a worst-case scenario rather than the best-case scenario of the first approach. In real trading the result would lie somewhere in between. The important point to take away, however, is that towards which of the two approaches the result in real trading would gravitate is dependent on where the limit order is in the queue.

 

Sophisticated market participants know with a reasonable degree of accuracy where their order sits in the queue. When their position is not favourable they can respond accordingly, either by cancelling unfilled orders, by ‘scratching’ trades with market orders, or by entering with market orders instead of limits where their position in the queue is unfavourable.

 

Do you know where in the queue your order is?

 

More on this to follow . . .

 

BlueHorseshoe

Edited August 13, 2012 by BlueHorseshoe