High frequency traders (HFTs) seem to be the evil redheaded stepchild of today’s markets. France implemented transaction taxes, Germany is pushing for registration, and the EU is proposing minimum holding times, increased pre-trade risk requirements, stiffer market maker liquidity requirements, mandatory kill switches, elimination of broker crossing networks, and the banning of tiered exchange pricing. All of these regulations are meant to increase market robustness, and to pressure high frequency firms to either provide greater liquidity, or to leave the markets entirely.
The macro goal of these rules seems to be to slow down the markets, decrease speculation and intermediation, reduce gaming, and curtail short-term equity trading profitability. If these rules are implemented, I am not sure that we will experience the perfect market that regulators are hoping for.
Do we think technology will slow down? Do we think the use of computers in trading will decrease? Do we think that by lengthening the time an order is in the market it will stop faster firms from having an advantage? Do we think that mandating a longer time-in-force for orders will make bid/offer size increase? Do we think we can force HFTs to become market makers if this isn’t in their best interests? Do we think that a market maker who doesn’t want to provide liquidity will do so because of a regulatory fiat? Do we think investors will want to pay higher commission rates or demand expensive capital from banks? Or that banks will suddenly (especially in cash-strapped Europe) open the capital window? Do we think new risk intermediaries will come into the market without reward, especially in an age of regulatory expansion? And do we think that just because we slow down the market, or place artificial market barriers, that gaming will end? Of course not.
Comments | Post a Comment
54 Comments to "The End Of High Frequency Trading As We Know It?":
ccwilliams
02 October 2012
Great damage has been done to the whole investment community by the arms race of HFT/algo and its domination of the exchanges. Credibility, jobs and concentration of wealth are all consequenses. If the broader world understood that the prime objective of large financial firms was to rip people off the game would be up. The markets functioned well without 10 milli second trades and quote stuffing I'm sure they will work again when it is properly regulated.
Comments (1)
Anonymous
02 October 2012
I agree, the proposed rule changes are only a first step in the process. What we need now is DIFFERENTIATION between market makers and other market participants. Market makers should be required to provide liquidity at the top of the market for some meaningful % of the time, and as a benefit for providing this essential market service, they should be rewarded with things like priority, or the ability to send in more quotes, or more frequent updates. We need to reintroduce the concept that a market maker is different from a customer, re-institute stricter quoting requirements for market makers, and provide a meaningful benefit for the service that the market makers will provide to the marketplace.
None of this however means that the initial steps taken to curb HFT aren't good ones, they're just not enough to solve the problem by themselves.
pchepucavage
02 October 2012
Seems like the markets reached their all time high long before the advent of hft-as did the number of ipo's and persons employed in the industry and I belive retail participation.Why therefore is it so necessary in view of the disruptions it causes.
Comments (11)
ltabb
02 October 2012
Its a really delicate issue. How do you provide enough incentives to market makers, without creating too many barriers to providing liquidity for the market (and I know folks will comment and say that these guys don't provide liquidity - and I agree - what they really provide is price discovery). But by placing a time in force rule all you will do is ensure that stale quotes (in which there will be plenty more) will get picked off. Now that said - the real problem is fragmentation and tick size. With fewer venues and wider ticks more liquidity will naturally pool at the wider tick and there will be fewer venues to queue jump. they are trying to take the easy road. Which I am not sure will work. I do think you are right about differentiating real market makers from non market maker HFTs that have no obligations and just free ride. But the benefits of non-market maker trading (lower costs and easier access) also accrue to investors - so we need to ensure that we don't mess those up as well.
Comments (303)
Anonymous
02 October 2012
I agree 100% that fragmentation and rick size are major problems as well....the problem is that changing those (in the right direction) doesn't seem to remotely be on the table. Right now, the SEC is continuing to allow exchange proliferation (see the Miami Options Exchange), and concurrently allowing sub-penny pricing to take place. So it just appears that we're on a path where these things will not change any time soon. If the SEC actually looked at the marketplace as a whole, and attempted to make some broader structural changes, that would be great. Instead it appears that we're going to get some "limitations" on HFT which will likely not work, and then hopefully we can see some alterations to the market making structure as well as dealing with fragmentation and tick size.
kurtkujawa
02 October 2012
Fragmentation and tick size is definitely the problem, for those of you who don’t remember, that was the road to hell that was paved with good intentions after the 87 crash. Larry you are correct that less venues and wider spreads will pool liquidity. Wider spreads and deeper markets is what will lead to true price discovery. Market Makers, HFT’s and tighter spreads with 100 share tick sizes DO NOT lead to price discovery. SIZE is what leads to price discovery; you can’t determine the right price based on 100 shares. 1. Wider spreads will keep the HFT’s out, just as it kept the SOES bandits out before them, because the risk was too great. 2. With wider spreads the whole rebate game is gone. 3. I am not sure why people think we even need market makers? Let me ask you; are there market makers at art auctions or car auctions or even on ebay? NO. We need to get back to the basics; the markets are there for NATURAL buyers and sellers. If someone really wants to buy or sell something, then take it to where it trades. We need to get away from the whole idea that if someone wants to B/S something, they have some type of right to execute it where they want right now. You have 2 options, 1. wait until someone wants the other side at your limit or 2. CHANGE YOUR LIMIT!
Comments (171)
steveg
02 October 2012
I think the point that Larry is trying to make is that some of the proposed solutions such as minimum resting times just wont work in the real world. I wrote about this myself the other week
http://fragmentation.fidessa.com/2012/09/28/is-the-time-up-for-hft/
The question of whether HFT is good or bad is a different one - but more important is how the industry moves forward given where's its at now
Comments (12)
kurtkujawa
02 October 2012
steveg,, “Whatever happens, it looks like this piece of legislation will cause yet another slew of unintended consequences which will doubtless be followed by more regulation. The real problem in all this, though, is that the current cycle we are in simply drives up complexity and costs for the market as a whole and this inefficiency means everyone suffers.” PERFECTLY STATED!!!!
Comments (171)
kennymcb
03 October 2012
The 87 crash - "How our child has grown since then!" -
Comments (49)
kurtkujawa
03 October 2012
child? more like monster!
Comments (171)
kennymcb
03 October 2012
87 the heady days of Magggies privatization BP 's listing being delayed etc. To me the biggest disrupt or was decimal inaction, as innocent and logical as it was - that really opened the flood gates the radical change.
Comments (49)
kennymcb
03 October 2012
*decimalization
Comments (49)
kurtkujawa
03 October 2012
decimalization and fragmentation were co-conspirators in the demise of the markets.
Comments (171)
kennymcb
03 October 2012
Agreed - the ripple effect of Decimalization was the one that nobody foresaw or expected, Fragmentation and the drive to "for profit" exchanges didn't really help ....everyone was then after a buck.....the act of investing was locked into short term exploitation of process anomalies.....bring back quills Blue Buttons and messengers carry gazillions of payments by hand!. who stop for a pint on the way and then sprint at 2.55pm
Comments (49)
maxima
03 October 2012
very true. there is no soft socialism. there are two possible states - either capitalism or communism. whatever public and left politicians elected by scared and poorly educated middle classes might think - everything in the middle of these two states - is just a transition period from one state to another. you cant have slightly regulated capitalism and you cant have unregulated communism. all those attempts to 'fix' capitalism are nonsense.
Comments (1)
shamlet76
03 October 2012
we do not have to preserve the HFT industry. it is a detriment to the purposes of the stock market, a capital outflow from the market.
Comments (149)
sarnuk
04 October 2012
Larry,
At some point I am hopeful we will get to the point where we call modern market making what it really is - HFT Scalping. The modern market maker model is all about a rebate strategy. Bid and get to the top of the book. Constantly monitor your position in the queue. Only get executed when you have a sufficient floor of other bidders underneath you. Get the rebate. If you perchance get executed when you are not confident of that floor underneath you.... exit. Quickly. Never lose. Do this in all stocks. "Add liquidity" when there is already liquidity underneath you. Don't play where there is not. It is the ultimately efficient short term scalping model, and unlike market makers in the traditional sense, these firms do not have customers - they are hyper-efficient short term prop traders and scalpers.
Now, they have every right to play and participate in our markets provide they do so on an even playing field. But why do we as an industry give them special status? Why do we create order types that enable queue jumping? Why do we bend over backward to give them (or any player) an edge? Why can they have exemptions from short sale rules? Why do they get incentives at all? If any thing, we should eliminate all incentives for all players that are divorced from true market supply and demand. That means eliminating payment for order flow- especially maker-taker pricing models. Might that make our order books look less deep? Perhaps. But the "liquidity" from these players is fleeting. It is there only when someone else backstops them. All in the know already understand that the "liquidity" you see is not what you get.
Ironically, I think if these incentives go away we may not even need to have discussions on appropriate speeds, cancellation fees, algo-mointoring, and kill switches, etc. The market will decide (and correctly decide) on an equilibrium that I dare say will be an improvement over what we have now. And systemically safer.
Oh the horror. Can you imagine a market place where buyers buy because they think a stock will go up, and sellers sell because of the inverse? Can you imagine a market place where the best-priced bids just plain match the best-priced sellers.
We know how for-profit exchanges and ATSs benefit from giveing these players every edge they can think up. But how does the public benefit? Is a (perhaps) miniscule spread reduction a proper compensation to investors for all the added baggae and systemic issues that this class of trader has brought to the market in recent years?
We have perverted the most elegantly simple process of matching real buyers and real sellers, and to the detriment and safety of the system.
Comments (160)
sarnuk
04 October 2012
Forgive my spelling.
Comments (160)
ltabb
04 October 2012
Sal,
I don't disagree with much of anything you said here. Well a few things. About incentives - the incentives in markets have been around for at least 32 years (the length of time I have been in the markets). IDBs on the fixed income side have always (at least from '80) charged the aggressor and the provider paid zero. That isn't an incentive but its pretty close. The incentives have been to populate my book with limits so the market orders come to my platform. That said - you are right incentives distort market behavior. If they were lifted, I am not sure much would happen, except the spreads would widen a tiny bit to make up for the lost rebate. And exchange market share may shift (which is why the exchanges pay rebates).
As for jockeying for priority - you are absolutely right and order types you are spot on. If you are first in the queue - why is there an order type that jumps ahead of you? That said, there are a number of order types that are excellent - hidden orders, floating orders, midpoint orders... these orders help folks get what they need done. As long as they don't step in front of others and go to the back of the queue - I am cool.
As for all of the strategies to only provide when I can flip. Not sure what to do about that but try to de-fragment the market. Limit lit, dark, and internalizers and pull the licenses of dead players, and once you hit a certain limit - make new guys buy the old guys licenses. or give folks a limited license so they have 1 or 2 years to get x% market share and if they don't - bye bye.
But fragmentation comes in many stages, not just venue fragmentation. there is price and time. As for price, I have been on record saying that we should widen tick sizes in the smaller end of the market, as for time - that is the one that is difficult as the whole technology and communications industry are fully dedicated to making computers and communications faster. I am not sure how you stand in the way of that. And if you try to put in artificial limits, orders will get stale and the most stale prices will get picked off. Remember SOES bandits picking off stale broker quotes. They were the scourge of the industry.
As for market makers - sure I don't have a problem if you change the incentive structure and provide market makers more benefits for providing more liquidity, and clamp down on non-market makers posing - however how do you distinguish the non-market makers from the investors. The same benefits that give non-market makers the ability to do hft is the same that makes is to inexpensive for investors to trade - so you need to be careful that you don't kill the goose... Also the level of regulation is already pretty significant, and I am pretty sure it ain't getting easier.
And as for a level playing field - I completely agree.
And as for a market for only buyers and sellers - I think you are living in some sort of utopia distopia. A market where there aren't speculators and where there are only natural buyers and sellers is a market where prices are all over the place, markets are not calm, there is little depth and only the top of the market has any real liquidity. Now that may be ok for the professional who has the sense to know exactly what they want to pay for an asset, however I am not sure that works for the majority of folks. Has there ever been a market without speculators? That would be some sort of auction market, where only naturals were able to play and folks would spend a lot of time waiting for the other side to show up.
Good luck with that.
Not that it isn't idyllic, but i am not sure it works in practice.
Comments (303)
sarnuk
04 October 2012
Larry,
First thanks again for providing this forum to debate, discuss, and learn from each other.
Regarding your last paragraph, I defintely think there is a place for speculators and investors of all speeds. And the market can decide whether their strategies work. Regardless of their speed, just please let them be speculators on return or direction of the stock - and not a rebate video game that is distortive. I did not mean to imply that we thought the market should just have long term investors.
Comments (160)
kurtkujawa
04 October 2012
Larry I agree that the incentives have been around for the last 30 years, and we all know how corrupt and worthless that was, whether it was the agency desks paying for order flow or the recapture firms that unfortunately we still have. As for spreads widening, that would be a good thing, as Sal stated previously. As for exchanges loosing share, that too would be a good thing, hopefully all but 1 or 2 would go away, then we would have deeper and more transparent markets.
As for the market maker aspect, I don’t think there should be any different rules for market makers vs. investors. I’ve been in the business for 25 years as a sell side trader, a market maker and now on the buy side. As far as I’m concerned there is no need for market makers. Everyone has direct access to the markets, so unless you want capitol, what do you need them for and if you do want capitol, the firm can trade that order then get out. If you want to keep the market maker concept, fine, I can live with it, but I still wouldn’t give them any type of preferential treatment.
As for a market of only buyers and sellers with no speculators? Once again you are distorting the definitions, every buyer and seller is a speculator, it’s just that HFT’s and to use your own words “Remember SOES bandits picking off stale broker quotes. They were the scourge of the industry.” Are one in the same, both hid under the guise of being market makers and pretended to add liquidity! As for having naturals only, the markets would be much calmer, since they wouldn’t constantly have HFT’s trying to get in-between every institutional order out there. The depth would actually be larger up and down the book, just as it was on the NYSE. You said yourself that liquidity would be pooled to one place, increasing size. As for it being “OK for the professional who has the sense to know exactly what they want to pay for an asset, however I am not sure that works for the majority of folks.” What the heck is that supposed to mean? Are you suggesting that we dumb down the markets for idiots? And what is wrong with an auction market? So what if folks waited around for the other side to show up? As I said before, as an investor you have 2 options, wait for your price or move your limit!
Comments (171)
johnh
04 October 2012
Many of the comments above are all well and good, but seem to overlook that we have a free market. Whilst there's no doubt demand amongst the traditional buy-side for the type of level playing field venue described above, the reason they're not directing all of their flow to such venues is because a lit level playing field doesn't exist, and probably doesn't exist in the dark either. The reason they don't exist is because as entirely level playing fields they are unable to attract sufficient liquidity from a single type of market participant. If investors trading with investors could generate sufficient liquidity exclusively between themselves, then the free market would provide them with the level playing field they want. But they can't; so venues offer incentives and rebates to other types of participant to create deeper order books.
Petition your regulators to wind the clocks back if you must, but beware of the unintended consequences. It seems somewhat ironic that the HFT phenomenon that is now causing so much consternation amongst investors and regulators is a direct and unintended consequence of MiFID and RegNMS. Yes, obviously there was competitive advantage to be had from being faster, smarter or both when trades were matched mostly on single venues, but the heady mix of competition for liquidity between venues and the resulting arb opportunities is what's fuelled such staggering HFT growth. Perhaps easy to say with hindsight, but this doesn't seem like an unintended consequence that was particularly difficult to predict.
Sure, HFT has an ugly side and nobody likes being gamed or queue jumped, but before we all start lobbying our politicians and regulators to regulate it out of existence, think long and hard about what it might be like trading off a dramatically thinner book with multi cent spreads and the higher fees that would result from less venues competing for your flow.
You can have one thing or another, but you can't have both.
Comments (2)
kurtkujawa
04 October 2012
Johnh, I have to disagree with you on the fact that the level playing field doesn’t exist because they can’t attract sufficient liquidity. That argument just isn't true. We have proof that when you have one market place (exchange) and less fragmentation, liquidity is deeper and more transparent. I hate to keep going back to the NYSE before it became just another electronic gateway, but it is undeniable that the markets were deep and liquid, true, the spreads were wider, but that was a good thing. Even though the spreads were wider, most stock traded between the spread, why, cuz the market was priced for size, which is why you had true price discovery, once the parameters were set by the institutions, the little guy could trade between and the smaller orders were matched up, they just weren’t displayed. So my point is, if we went back to a CLOB where the displayed market would be the accumulative size to get to say 25k shares a side, yes you would have a wider spread but it would be sizable. stock would still trade in between for the small orders, but if an institution wanted to B/S size they could pay the price of the spread and get a start. Someone please tell me how that is a bad thing and who is hurt by that, other than the pay for play and the for profit exchanges.
Comments (171)
sarnuk
04 October 2012
Johnh, I will just add this respectfully.
Do you really think what we have now is an unintended consequence, or is it a well-executed plan? Who wrote REG NMS? Did the industry write it, re write it, lobby, delay it, lobby more, delay, and finally implement (2004-late 2007)? In that time the industry was busy at work figuring out the game. How lucky for them that they coincidentally happened to win unintentially.
Comments (160)
kurtkujawa
04 October 2012
LOL
Comments (171)
johnh
04 October 2012
Kurt, In the 'good old days' that you're referring to you had floor traders or locals. They could see the flow and smell the fear. They got in front of you, gamed you, had all manner of devious means of finding out where your stops were and generally played havoc with your orders. But they also provided liquidity and were fun to go drinking with too. You effectively paid them to provide liquidity by giving up a little bit of each trade. When they moved upstairs and became click traders, not all of them made the transition and many ended up driving taxis, but those that did became adept at spotting you in the order book, just as they spotted you on the floor. Again, they earned their living off you not being quite as quick as them but they provided liquidity. There weren't too many people shedding a tear as they started to get beaten at their own game by the machines and more and more of them ended up driving taxis with their mates from the floor. My question to you is, how is that liquidity going to be replaced if we now regulate HFT out of existence?
Comments (2)
pskopp
04 October 2012
Kurt> true, the spreads were wider, but that was a good thing.
No - it was not a good thing except for the human market makers.
Kurt> Someone please tell me how that is a bad thing and who is hurt by that, other than the pay for play and the for profit exchanges.
Ummm.... how about retail investors placing market orders?
Kurt> Are you suggesting that we dumb down the markets for idiots?
You don't care about retail and don't want to "dumb down the markets for" them... I would guess most HFT participants would agree... ideally markets should not be dumbed down for anyone, including ludite institutions.
From a regulatory standpoint however, it is the little guy (retail trader with small size one and done market orders) that needs protecting - not the professional and institutional traders who find their technology is outgunned.
Comments (28)
kurtkujawa
04 October 2012
I will tell you this, what you say is way over blown, yes there were some bad eggs and front running jessies, but buy no means was it as prevelant as it is today. I will also say, i would much rather see a 25k share market and know who is trying to steal from me than have 100 share markets and only have an idea who is stealing from me. My point is i dont't think we need market makers or specialists, I think that a CLOB would satisfy size and depth while still letting odd lots trade between the spread as they do in the dark now, we just need to get rid of the fragmentation, which inturn would get rid of the majority of HFT since there would be no latency arb or rebate models. And yes, alot of them were fun to have a drink with and yes most are driving taxis now!
Comments (171)
kurtkujawa
04 October 2012
pskop,
Sorry but i have to disagree, wider and deeper markets are a good thing. As an intitution, nobody wants to trade in 100 lots, the more trades the more leakage, if i can take an offer for 25 to 50k and not move the market that is a good thing, and it is a good thing for all the small retail oredrs that might have been on the book at a lower limit, but traded at the higher price because i came in and bot the book up a .25, all those orders below that price got the .25. So your retail guy actually got a better price.
We agree that the market should not be dumbed down, it was Larry who suggested that.
From the regullatory standpoint, just an FYI, it was the SOES and now the HFT's that hide behind the retail or little guy rules. Reg MS and MIFID were implimated for the so called little guy, the only problem was the lobiests pretending to represent them were nothing more then the SOES firms usiong the term as a cloak. That my friend is what happens when you pick sides and try to protect one person from the others. If you want FAIR and LEVEL markets, then you need the same rules for everyone, regardless if your retail, institution or scalper. When you write different rules for different players you create more fragmentation and allow ways to game the rules. NOBODY OR GROUP should have special rules.
Comments (171)
crammond1964
05 October 2012
sadly larry volumes and confidence has disappeared from our markets . HFT and ATS have had time to improve our markets and their performance is now being judged . We had to leave floor trading behind because it was hindering our mkts performance and now HFT is suffering the same fate .
In years to come this decade of HFT and colocation and under regulation will make interesting reading . Our markets can and will succeed without them .
Comments (252)
kennymcb
05 October 2012
That sad fact is that long term investing, financial trading strategies have been reduced to a scientific project, getting as close to or matching the speed of light.
Trading strategies in essence are being dumbed down - The analogy would be the olympics being scaled down to a mere 9 seconds of activities, instead of 2 three weeks, as the 10 meters was the only important event. .... The market has lost itself in the execution process - how much money could be saved if firms actually leveraged modern technology to really improve their Middle and Back office legacies?
Comments (49)
canelson
05 October 2012
I saw a suggestion a few weeks ago that I’m surprised has not gotten more traction or air-play: that the markets be put on a standardized clock and execute once per some defined interval, in a coordinated way across all venues just as the electric grid is synchronized. In effect it would become a series of call auctions, which are a great way of creating both price- and size- discovery simultaneously. The interval would not need to be long in human term, perhaps a half or even a tenth of a second, but it would be ages longer than the microsecond timeframes that we’re dealing with now. The technical challenge of synchronizing exchange clock-ticks would be a doddle compared to the feats of engineering that are already taking place. Under that rule, there would be no need to regulate the frequency of order placement or cancelation – orders could be placed and canceled at any frequency, but trades would take place on the schedule. Furthermore, the advantage of speed alone would be eliminated by changing the allocation of executing shares from price-TIME to price-SIZE. This would naturally encourage people to place larger orders. (How that would affect the retail 100-share order, I admit I haven’t fully worked out.)
Still, in all these discussions of mandating wider spreads and longer order-lives, it seems to me that this idea, whose clever suggestor has no doubt thought through more than I have, warrants further consideration.
Comments (1)
pchepucavage
05 October 2012
Here are comments from two clients who would like to hear about the short side of hft
Having served as head of ops and admin for a $60 billion buy side investment group that managed only pension funds, the concept of buying 10 stock every day where 6 are the result of a short sale that requires a borrow to make delivery, seems to be a source of value dilution, governance stress (setting aside the continuous vote tussles with the custodians, we regularly wrote 7 digit checks because the custodian could not tell us how many stock we actually owned or had been delivered so that we could sell out of a position the client had directed us to and by the time the custodian sorted it out the value of the stock had declined – its call a make the client whole payment), and benefit to someone other than the buyer.
In the wisdom of regulation, attached are the historical rulings on what a market maker is. How does this relate to high frequency trading strategies where computer systems ping off other computer systems to determine if they are real buyers and sellers? What I hear from high frequency traders are strategies, which are in theory not allowed under market maker exemptions. They are not playing a neutral party between buyers and sellers. How does this change the concept of market making? What about high frequency traders that are short day after day 60% of their trading volume, is that market making? If market makers are to be flat at the end of each day, where does the profit lie splitting fractions of a penny that all these high frequency traders at the Roundtable, such as Citadel, talk about if it relates to pure regulated market making?
I think we have moved into a new era of trading without a change in the regulations of market makers. This is somehow surpassed the SEC’s purview and they seem to be considering high frequency trading as a liquidity process that provides something to the marketplace that market makers normally would. However, these systems are far from neutral parties making fair and orderly markets between other participants, but rather are predatory systems for profit, taking positions that are not neutral. Should a market maker exemption exist at all for high frequency trading strategies?
Comments (11)
kurtkujawa
05 October 2012
one word NO
Comments (171)
dwgerhart
05 October 2012
The HFT problem is one of noise at the edge of the market. The policy concern should focus on the impact if any this noise has on the more econmically important matter of executing big trades. In that regard I think the HFT has further exposed the limits of our traditional market structures to deliver significant market depth and meaningful pricing for large positions.
Comments (17)
kurtkujawa
05 October 2012
HFT has not exposed the limits of our traditional market structure to deliver market depth and pricing for large positions, HFT is one of the problems causing that lack of depth. I have been trading institutionally for 25 years, the problem is that the traditional call auction market was highjacked first by the SOES bandits and now by HFT which takes advantage of Reg MS and MIFID. The road to hell was paved with theese good intentions after a knee jerk reaction by an inept SEC and a congress that knew and knows knothing as to how, why, markets exist. Once 100 share markets became the norm, and multiple for profit exchanges evolved, large orders had to be split into small orders and spread all over in order to try and hide from the preditory algo's. The biggest misnomer perpetrated to some unthinking buy side traders back in the early 90's was that tighter spreads with 100 share markets would actually reduce trading costs because some slick TCA provider told them so. I warned many of them at the time and now the chickens have come home to roost.
Comments (171)
kennymcb
05 October 2012
We are all clucked
Comments (49)
anonymole
05 October 2012
Do we really have a problem with HFT? We may have issues with the algorithmic mechanization of the markets and the cascading chain reactions they seem to propagate; against which we obviously need protection. But all the rest of this? I wonder...
As a retail "investor" (not a day trader or millisecond trader - an INVESTOR) can I buy and sell at a price that is near my target? Within a nickel or so? Absolutely. I'm an investor after all, I intent to invest my monies in equities that I predict and want growth within. I'm in it for the long haul, many months if not years. I get my price - I'm happy.
As an institutional "investor" (once again) can I get VWAP, or TWAP for the 500k shares I need to move? And can I actually get such a trade done in a day (or so)? Absolutely. Yeah it'd be nice if I could shave a penny or three off of my execution algo price, but I'm moving tens or hundreds of millions of dollars, I just need efficiency getting there.
What do either of the above care if the gerbils driving the microsecond speed wheels in the machine get ground to dust? Let them eat each other alive for all we care. But yeah, don't let the gears jam up or let the machine slam to a halt. If HFT is a way to feed the gerbils then by all means... And all this talk of reduced volume in the equities market being caused by lack in trust due to HFT and the like? Yeah, it sure looks like it's happening but I think for apparent reasons undiscussed. Exhibit A: The Baby Boomer drive to low risk investment.
http://www.schaeffersresearch.com/images/commentary/2012/121004spx3.gif
Look at the way the boomers ate up (and then dumped) the tech bubble. No qualms about diving in then. Maybe the reduction in equity volume is due to the boomer's general redirection in risk appetite? They are, after all, getting older and older. Using this drop in volume as justification for the vilification of HFT (not that I don't despise investment banks as much or more than the next guy) seems like a snipe hunt.
Comments (83)
kurtkujawa
05 October 2012
As for volumes, i agree with you on the baby boomer thing as far as moving money around, no question about that. As for HFT, no offense, but VWAO, TWAP is no way to trade, not to mention the fact that those averages get distorted wheb the preditorial HFT's are runnung the stock. So even if you trade like a lemming and just want to be average, the average is skewed. over the coaurse of a year and 100's of trades, thoses so called gerbils are costing your pension and 401 k millions of dollars. Personally I woul;dn't call tha gerbil food.
Comments (171)
shamlet76
05 October 2012
i tried to sell a stock yesterday. this stock is on buyout at $0.27. the ask was $0.27 and the bid was $0.267. i put a limit of $0.267 order to sell. as soon as i executed the order, the ask went down to $0.267 and the bid of $0.267 disappeared and suddenly dropped to $0.264. there was no way a trader was watching this low volume stock constantly - it was a puterized trader who had software that was trying to take my $. i cancelled the order. and i was the only one who wanted to trade yesterday.
as it happened, FINRA had called me about market structure issues a few minutes later so i told them about my experience. an associate director encouraged me to make a complaint about it so they can look at the way that puter is programmed, the puter's trading pattern. i think this is an algorithm trader, not necessarily a HFT trading system, but it's all the same to me. so i made the complaint.
Comments (149)
ltabb
05 October 2012
I have a quick question for this community.
As part of my CFTC HFT Subcommittee responsibilities we are trying to put the HFT definition into a broader context of algorithmic execution.
Here is the question. Do most folks think of HFT as a subset of Black Box execution and Algorithmic as buy-side / human generated electronic execution or do most folks think Algo includes HFT.
In that case is there a way to separate traditional buy-side algos from hft algos using more descriptive language so that this whole nomenclature is less confusing?
Thanks for your help
Comments (303)
ltabb
05 October 2012
Shamlet,
The problem is you put in a limit. A limit means that you need a market order to execute against. If there was no market order, a limit just will sit there. In addition, when you place that limit, other folks - both computerized and non-computerized can move their limits around to either be more or less competitive.
If you really wanted to trade, you should have used a market order.
That said, depending upon how large your order was, and how it was routed, you may have gotten a bad execution if someone had sniffed out our market order and changed their limits if your order was for larger than the limit you first hit.
Comments (303)
shamlet76
05 October 2012
in my opinion, the HFT order cancellation rate sets it apart from algorithm trading.
i think there is a problem with algorithm trading also, in that some puters are programmed to execute illegal strategies.
this is what i feel, as an observer of the stock market.
India ETFs in Focus After Flash Crash
Comments (149)
kurtkujawa
05 October 2012
Larry,
Not all algo's are HFT, yet I would say all HFT use algo's. Also, believe it or not, I know that not all HFT is preditory. However I stand by my opionion that all HFT distorts and creates much more volatility in the markets.
Comments (171)
anonymole
05 October 2012
Predatory market taking = HFT
Contributory market making = !HFT
Unfortunately "HFT" is a poor acronym to describe the predatory market taking behavior. PMT-HFT then should be the villain so accused, versus the CMM-HFT benign version. And yeah, if the PMT-HFT gerbils grow teeth too large for the bodies and start preying on true investors - a means to trim their incisors back should be adopted.
Comments (83)
shamlet76
05 October 2012
larry, the problem is not with my limit order. the problem is with the fact that someone had a puter that was trying to give me a worse price as a strategy. they were displaying $0.267 pricing to lure me to sell at that price. they had no intention of executing at that price. and with this particular stock, if i wait a week or two, i will get $0.27. so i decided to wait because i was unwilling to let their puter take my $.
and if i had marked the sell order as a market order, would i have gotten execution at $0.19? or $0.04?
Comments (149)
sarnuk
05 October 2012
I think high frequency trading must have some element of holding period. The Boston Mutual Fund buying 500,000 shares of IBM in a vwap algo is using technology - true. That fund is using colocated speed (indirectly) through their broker - true. Does that make them HFT? Are they in the same bucket as an automated flipping proprietary trading firm firm flickering all day long, playing rebate arbitrage, or "get me a risk free shot at an exchange rebate", and going home flat or close to it? I know that the latter wants very badly to be defined alongside the former, for obvious reasons. But if the goal it to define HFT and we knowingly allow them to be placed in the same bucket, we are doing the investing public a grave disservice.
Comments (160)
sarnuk
05 October 2012
anonymole... an entire cottage industry that nowadyas call themselves market makers have spent a good 5 years trying to present themselves to the public, the buyside, and the regulators, as "market makers" that "provide liquidity and shrink spreads". Two things.
1) Since 2007, when modern Ultra-HFT took off, spreads are nearly the same. I suppose in the largest 1,000 stocks you can maybe...maybe make the case that they dropped the average spread 1/2 a penny. I have seen studies and charts, including one provided by RGM to Bloomberg for the tv spot I did in August, that have shown that spreads have actually widened since then. In other words my point is that general technology adoption and efficiency is what brought spreads from 25 cents to 2 cents - not "HFT". Since they amped up the market spreads have widened.
2) These CMMs you refer to (contributary market making) are playing a game. They have automated with great speed an algorithmic hunting program that has them bid and offer when (1) they are near the top of the limit order book queue AND (2) there is depth behind them to stop them out flat. Worst case they collect a 33 mil rebate when they buy, and can sell at the same price for a cheaper take fee than their rebate. Risk free. Nice. And we allow this.
They may have fancy websites, and they may keep trying to repeat over and over thatthey add liquidity and provide liquidity, but every time they say that please think of their algo I mention in (2) and put their words in context.
They "add liquidity" when there is already liquidity there.
This opnion is mine and mine alone. Your mileage may vary.
Comments (160)
sarnuk
05 October 2012
An algorithm is not writing my repsonses.... my spelling and grammar is proof.
Comments (160)
kurtkujawa
05 October 2012
I would like to add that not only do they want to pretend they are the same as Boston MF, they also want to pretend that they are the same as Broker NY market maker in order to get trading rule exemptions, some even want to pretend they are exchanges for similar reasons. If you want a level playing field, level the field, NO PREFERENTIAL TRADING RULES, regardless if you are an institution, retail, MM or whatever catigory might be invented in the future.
Comments (171)
dwgerhart
05 October 2012
I believe that connecting HFT and Algorithmic trading tightly is not helpful. The definition of HFT starts with the definition of capability rather than application. HFT becomes possible when two things exist, the first is automated market places with the ability to accept and act upon order instructions in small time intervals and at high sustained transaction rates, and the second is the ability for market partipants the generation order instruction at similar rates. The definition of a small time interval varies over time, for arguments sake lets define the threshold today as under 1 msec.
To generate or process order instruction in a 1 msec time frame requires computers and computers must be instructed. So the next question is when does a computer program become an "algorithm". Is chopping 100k share order into 100 share pieces and executing them at best prevailing market prices constitute an "algorithm" for the purposes of this discussion. If it does thats fine but its not helpful because this algorith has been in operation for decades with and without computer support. Is trading against the small price discrepencies that might exist between two market places an "algorithm", again this has been going on for a long time.
We face a problem when the interfaces to our marketplaces can accept order instructions and act upon them faster than market participants can interpret the results and make economically rational reactions. When this occurs we simply overwhelm the liquidity managemement capabilities of the market place.
Bugs in a very simple computer program can cause a flood of orders to be generated which can overwhelm the liquidity and risk management capabilities of either a market place or a participant. It is perhaps the risk management systems of the market places or the participants which should be throttling these flows and not the trading and matching algorithms them selves.
Comments (17)
kennymcb
05 October 2012
From my very simplistic perspective - Buy side algo's provide assistance in execution. HFT's are parasitical in nature, by that they need an already made market to exist. Exposing yourself to the market is to expose yourself to risk, HFT's by design are created to live off others risk appetites, by gaming a system to generate revenue from rebates seems. HFT's are like Pilot fish feeding off a sharks teeth - sometimes they get bit - the problem is the world has too many Pilot fish currently.
Comments (49)
pskopp
05 October 2012
Larry,
I think it is too late to come up with a definition for HFT. The more people know about automated trading, the more questions they will have about a simple boolean characterization.
It is a lost cause when terms like HFT, Alg, Quant, Flash, Black Box and Cheetah are all used to describe a vast swath of the trading eco-system.
There are thousands of strategies and they fall along a continium of dimensions including average holding period, maker vs taker ratios, hedged or naked, single or cross asset, types of signals processed, market impact etc.....
Any attempt to come up with a definition for HFT (regardless of what that definition is then used for) is sure to be rejected by most participants as flawed, even though there will be countless different and conflicting views on just what those "flaws" are.
If the point of CFTC looking into HFT is to attempt to regulate or investigate a subset of traders or strategies, then the focus should instead be on identifying what problems are trying to be solved, and then come up with a new term for any trader/strategy that creates (or might create) the problem...
If high quote to execution rates is viewed as a problem, then coin some new term for high quoters. But trying to define a term that alread means different things to practioners, regulators, financial writers and bloggers is really a lost cause.
Comments (28)
ltabb
05 October 2012
guys - thanks for the help. As you see it is not so easy. not that, that was my point, to show you how hard it is to come up with a good definition. But that is our task. Thanks again.
Comments (303)