Maybe you could do it for less than 400K. I couldn't. You know how to program at a high-level. I don't. That would take me years. You're talking buying servers, co-locating them on some exchange, then developing, writing, and evolving some piece of code to compete with the likes of Goldman. Okay, man. I'm not PI, over here. lol Let's be reasonable. Reasonable says a layman trader (like me) who has no programming aptitude would have to hire somebody (like you!) to do it for me. That's gonna cost $$$$. That's not doable for low-cap traders, or funds. That's my point on the development cost-side. As far as HFT, it's not all front-running. But a lot of it is. Anything that polls market orders and denies ghost orders get hit, is total bs that gives a Lock Trade to anyone with the skills and coin to implement. It's not rocket science. Goldman polls both sides (or sold the information from the exchange and flashed prior), and trades the heavy side = front running. I agree ghost orders have to be executable and flash trading, gone. The tools change, but never the game. This stuff is as old as the hills.
This is the kind of hostility that makes ET a waste of time. English comprehension is not an issue achilles28... and look at the second quote... you said they don't save money they do limit market impact... then you say those two things are the same thing... so they do save money, right? Confused, convoluted thinking... contradiction... mixed in with a big dose of anger and hostility = a pretty typical ET poster. <sigh> First of all, I owe the ET community an apology. This happens every couple of years now... I read a few posts and then write a little flood on the topic, then stop posting. In this case, I had a discussion re the public perception of HFT / non displayed liquidity with a friend who suggested I take a look at this ET thread... ugh... mistake. Sorry and sorry to have posted some confrontational messages late last night. Give me the liberty to do two things in this post and then I will slip back into lurking mode for several years lol. Let me address a couple issues remaining with this discussion and then some general words on why I didn't find ET to be a helpful part of my evolution as a trader. First of all, there are still a lot of misconceptions on this thread about what HFT is, what frontrunning is etc. Just because I know someone is buying and buy in front of that order, that is NOT frontrunning. It depends on how I got that information. If I got it from looking at a chart formation and realizing that a lot of funds are going to need to buy a stock, or from fundamental analysis that tells me a stock is deeply undervalued and people are going to have to buy no one would say that's front running, right? Well it's the same intraday... if I see that people are going to have to buy intraday because of a pattern on the chart or in the order book and I take a position in front of that order, that's still not called front running. It's called "trading" and it's what we do. A lot of trading involves figuring out where buyers and sellers are and what they're going to have to do. That's what trading is. If I do it by probing order books (whoever said you can probe market orders is wrong... you can probe limits) either by hand or with an algorithm, that's also not front running. If I'm standing on the floor of an exchange and I know a broker has a big order because I can read him and tell when he's nervous about filling a big order... and I take a position in front of him... that's also not trading. it's good trading and while it might cost the buyer a higher price here, on balance this kind of activity adds liquidity and tightens spreads. it's also not free money for me because there is some risk that a smart seller could be tricking me. maybe he's making it look like he's buying when he's actually selling. I used to do it all the time in thin stocks... say it's 10.25 @ 10.60 and I have a ton to sell... I would BID .28 and algos would penny me at .29... i would bid .32 and they would penny me at .33 etc... until it's bid in low 40's at which point I hit the bid and sell to the unwitting high frequency trader who just got screwed. Bad trades happen to black box / HFT algos and the failure rate for new programs like this is extraordinarily high. It's not free money that Wall Street is taking from poor unsuspecting retirees trading Ameritrade accounts. It's called "trading". On the other hand, if I have knowledge of the order because I am a broker or because someone told me someone had to buy, and I personally take a position in front of that otder, that is clearly front running and is illegal. I believe that happens very rarely now and offenders are punished when they're caught. That's just not what is going on with HFT. Re the cost of Dark Pools, the poster who mentioned fees is correct. The customer pays a lot for Dark Pool access which is why it's profitable for the banks running those pools fyi. However, for me the customer they often ended up being cheaper. Any large trader should understand transaction cost analysis (Plexus Group is one of the leaders in this work) which shows consistently that market impact is usually a bigger part of the cost than commission. So the answer to which is cheaper is, it depends. For you doing 1000 shares of XOM in your Ameritrade account, obviously the primary market is cheaper... for an institution doing 1,000,000 shares the question is more complicated and depends on how quickly you want the order done, among other things. Part of the job of an institutional trader is knowing which venue is ultimately cheapest for certain orders. You won't always be right, but you will be right more often than you're wrong. Again, it's called "trading". I find achilles28's defeatist attitude toward algos and HFT discouraging. It's sad to see someone critical of those who are successful, but that seems to be the predominant attitude on ET. Big traders generally got big because they were good at what they did, so they got more money and hired more smart people. If you don't have the tools to do something you want to do, get those tools. If you want to do HFT, then learn to program. If you want to trade patterns, then learn statistics, etc etc etc. If you want to learn to trade, find a mentor with a style that matches what you would like to do. That's the American way right? If you want something, find a way and figure it out. My journey is proof that this is possible--you really can start out with a very small amount of money and eventually make it in this business. As I walk past the NYSE on the way to my office most mornings, I often reflect on my first idiotic trades, my losses, the times I almost quit... and now it's all worth it and there's no reason anyone else can't do the same thing. But my caution to you is that it is probably a lot harder and longer road than you would expect. you are setting out to do something enormously difficult... something that goes against the way your brain is hard wired from millions of years of evolution. it involves sacrifice and expense... don't be afraid to pay for information and education (but i also have had contact with numerous "market wizards" and have been very surprised at their generosity and willingness to help someone out.) There are definitely things I wasn't able to do in life while I was learning to trade... it takes completely focus to the point of obsession, an understanding family, and the right personality to persevere... and frankly it probably doesn't hurt to be a little too stubborn and too stupid to quit. I turned down jobs with bulge bracket banks that would have made a very comfortable living for me and my family for the uncertainty of trading income. in a way though, this is no different than the sacrifice and uncertainty you would assume opening a restaurant, or a sign store, or any other business... or developing a skill such as being a professional dancer or athlete. sacrifice, time and hard work. but... be clear... this is a business and will only reward a real commitment and here's where we arrive at my point: the problem with ET is that most people here are not successful. Successful traders generally do not waste their time posting here. There are people here who make outrageous claims and who try to sell some product and the community reacts with overwhelming negativity... which is justified but it becomes an overwhelmingly negative attitude towards everything. Leaving ET and the old Yahoo groups was a milestone in my development as a trader... it was really holding me back and if you're serious about trading you might want to think about the issue seriously. Sorry this became so long... back to lurking mode now.
Now that is a true trader-that-need-to-dig-into-exchange-rulebook talking. My advice is not to talk about HFTs, dark pools when you don't have basic understanding of order matching. On flash orders like it is described, it doesn't seem like front-running because the HFT algo is on the other side of the flashed order. Front running is more like removing sizes in front of an order that will likely remove more, move the market and enable you to exit higher/lower. This practice of flash is unfair because there are guys waiting in the queue to get a piece of orders but it's not front-running...It's more like "queue-snaking". Talon, thanks the explanation of dark pools and HFT. We can disagree, but you are for sure a patient guy willing to help and explain.
I cant believe the amount of bad information going off in this thread. Talontrader, and Momo are the only 2 who seem to know what they are talking about in regards to dark pools. There is absolutely no difference between a darkpool and a hidden NSDQ order, except that if you hide on NSDQ you get a credit. It used to be millenium could print some of their volume as arca prints, and just not print alot of it, but now they even show you their damn hand. Lets say the bid on a stock is 30.57 if you start seeing prints go off at 30.573 it means there is a millenium buy order there. If a stock is offered 30.57 and you see prints going off at 30.567 it means there is a millenium seller there, the large majority of the time millenium prints this way now so they are not even a dark pool anymore. If you see a stock bid 30.57 offered 30.58 and you see prints going off at 30.575 it means either a crossfinder buyer or seller is there. If people would do the slightest amount of work educating themselves on this they would not be bitching about it, the only people bitching about it are people who really dont know how the market works, and dont take the time to try to figure it out, How bout when you see prints going off at strange numbers you start testing orders to figure out what ECN/darkbook it is at? Or do what talon mentioned above, and simply test all the darkbooks that are out there, or atleast the 2 big ones i have mentioned, if you test and see a darkbook order following a stock arround wherever the NBBO goes odds are the guy has a lot of stock to sell/buy. I made close to a mill for my company raping these dark pool orders and a few of the traders i associate with probably made over 5 off of them, Swift trade as a company made well over 50 million gaming these guys in these dark pools. If you dont like the dark pools then why dont you figure out how to game them? Although i willl warn you what we were able to do to them is basically dead, as almost every PROFESSIONAL now checks for them to see if they are there. All the dark pools are is another ECN, anyone with trading software has access to the 2 biggest ones millenium and crossfinder. In all of my time trading against these dark pools i never once caught an out of market print it always had to be NBBO. 95% of software has access to the 2 biggest darkpools the ones i mentioned above so i cant even fathom that anyone who actually trades for a living would come on here bitching about them.
Hi talontrading, Thanks a lot for your explanations. I am also in NYC, if you ever want to go out for a beer on Friday evening shoot me pm. Would like to meet you in person. Regards, redduke
No, a true trader doesn't need to know the rules of an exchange. If you are a broker and rely on one of those certificates (series 9, 13, 15, 17, 75, 89......), yes, you need to know. After you know all the rules of all the exchanges of the world (Do you know the rules of Zhengzhou Stock Exchange? I do, one of their rules is that every stock trader must be a communist party member and read Chairman Mao's slogans before market opens. One of the slogans reads: The US imperialist is a paper tiger.), you still have no clue about trading. Let me reiterate my point: you may know all the rules of all the exchanges, but you still don't know how to trade. You may know all the details of dark pools, swimming pools, garden tubs, antique cast-iron kitchen sink, but you still don't know how to trade. oh, keep your advice to yourself, nobody wants it. ps: share what you know about the exchange rules with Madoff, he was the chairman of Nasdaq. But you will surely be disappointed if Madoff tells you this: I haven't traded since 1972, I have been busy with the exchange rules.
No, you said they don't save money but minimize market impact! That's what the quotes (" ") are for. Because that's the gist of what you said. Intellectual honesty isn't a strength, apparently. But whatever. Oh brother. Please, man. You clearly can't distinguish between what another person says, and what they attribute to you, via basic english. And since you did say it, perhaps you should own up and call a duck, a duck? And the hostility started with you, btw. You made a few jabs at my equity exchange/regulation knowledge. So, I returned in kind. Boo-hoo. Your entire thesis to rationalize HFT is purely semantical. Not coincidentally, as some or most of your trading income is derived from it... To paraphrase (you) for a moment, when a broker knows someone "has" to buy, and trades ahead, its bad. When a trader probes with fake volume (or gets flashed), and trades ahead, its 'legit' (sorry I used quotes there. Don't mean to confuse anyone). Flash orders from the exchange to a Broker/Dealer is not just an inkling of who might trade and when, but THAT REAL TRADERS OF COMMITTED VOLUME NOT ONLY WANT TO TRADE, BUT HAVE ALREADY SUBMITTED HARD EXECUTION ORDERS TO THE EXCHANGE FOR FILL. AND THEREFORE, "MUST" TRADE, RIGHT NOW! So when a Broker Dealer gets pending market orders flashed before those orders get executed, that's 100% hard-and-fast insider info, on who's buying how much (and when), and who's selling how much (and when). That's 100% front-running, by your definition. Plain and simple. You're obviously a smart guy. But totally fooling yourself, on that. As for the parting jabs about my apparent non-success. If you're a straight-up algo-dude, I could out-trade you any day of the week. Discretionary traders catch waves. Algo's get the leaks. Sorry, but you lose on that one. Have fun lurking.
my only complaint about them (other than when i'm in a trade and i want them out of the way!) is what steelers said, when there's prints crossing at a teeny price that i don't have access to, and can't put a bid/offer out to get above/below them.
...and the traditional media is full of them. Just listen to the 'File on 4' on high frequency trading (a BBC Radio 4 programme from 03/11/2009, available in their audio archive at http://www.bbc.co.uk/programmes/b00nk55r ) Now here is an unbiased piece of media reporting for you! Not a single buy-side representative (pardon - trader) was interviewed for the entire 40-minute duration of the programme. Instead we get an ECN operator, whose extremely low-latency high-throughput technology gets easily mixed up in listeners' minds (by an 'unbiased competitor', i.e. a traditional exchange representative speaking of 'nanoseconds' - really) with the actual trading speeds of algorithmic traders, some of whom manage to be competitive with half a second latency (so I'm told. Then we hear some unfounded allegations of front-running from yet another sell-side representative - a sleepy institutional broker whose dinner gets increasingly eaten by modern fully automated traders. He even uses the analogy of smoking being good for you (as HF traders are allegedly good for the market). Which they are, because they provide liquidity making spreads narrower even by a sub-penny just to earn their daily wages (so I'm told. Or lose 5-times that falling into liquidity holes created by those 10 AM fat-finger traders (I mean these guys pose a real threat to market stability - they should be forced to trade VWAP if 1989-style crashes are to be avoided). And when it comes to dark pools, guess if we are allowed to hear from an actual mutual fund trader, who could just state the obvious of lowering market impact (stopping other traders from backing away from a large-sized order) by making size hidden, which results in improved (or at least not worsened) execution prices for the fund shareholders. I mean, such splendidly unbiased reporting, quite distinct from the usual red-tops jumping on the bankers-blame bandwagon...