Readers who frequent the comments section have been treated to a spirited discussion over the past couple days surrounding the arrest of Nav Sarao, a futures trader accused of manipulating S&P 500 E-mini futures and acting as a "significant factor" in the Flash Crash of May 2010.
Mr. Sarao is a bloke who trades from his bedroom at his parents' house near Heathrow Airport.
The suggestion is that Mr. Sarao layered a number of 600-lot offers in E-minis, which apparently scared the bejeezus out of the financial world and forced a huge number of market participants to simultaneously panic, selling stocks with literally reckless abandon. For masterminding this financial maelstrom, he cleared less than a million bucks.
By Macro Man's reckoning, that makes him either the worst evil genius or the biggest patsy we've seen in quite some time.
This is not, of course, to condone the sort of behaviour in which Mr. Sarao was allegedly engaging. Spamming phony orders is a reprehensible practice, and as a punter there is nothing more frustrating than trying to pay an offer that you see on screen only for it to magically disappear the moment you click "Buy."
However, given the prevalence of similar and even more morally dubious practices in single-stock world, it is bemusing to say the least that the only entity to be arrested (to Macro Man's knowledge, at least) is a small-fry punter who conveniently happens to be a foreign national. No mention at all has been made of the HFT algos, for whom spamming phony orders and front-running is a (very profitable) way of life.
When it comes to the Matrix, it appears that the regulators have thus far decided to take the blue pill.
Macro Man is hardly alone in his skepticism. He has unsurprisingly been chatting offline with a few mates with many years of experience at the coal face of financial risk taking, and no one has thought the situatio0n anything short of absurd. In the context of one of these conversations, one friend offered a suggestion on how to replace the badly-broken market mechanism with a few simple rules to replace many of the current Byzantine regulations:
1) If you are acting as an agent and an order for a client gets filled (full or partial), then the client gets filled at the execution price- no front-running allowed.
2) You cannot submit any orders for more size than your margin currently supports.
3) Any order submitted to an exchange or market must remain valid for a non-quantum length of time (call it one second) before it may be cancelled.
4) If you buy or sell something, you must make or take delivery within the specified amount of time. Failure to do so should incur a hefty financial penalty (scalable for repeat offenders) and a ban from further trading until delivery is made/taken.
To this, Macro Man added a fifth:
5) All participants trading on an exchange should incur the same exchange fees. Exceptions can be made for locals trading in open outcry pits.
Macro Man is curious what readers make of these suggestions and whether they have a few good ones of their own....
Mr. Sarao is a bloke who trades from his bedroom at his parents' house near Heathrow Airport.
The suggestion is that Mr. Sarao layered a number of 600-lot offers in E-minis, which apparently scared the bejeezus out of the financial world and forced a huge number of market participants to simultaneously panic, selling stocks with literally reckless abandon. For masterminding this financial maelstrom, he cleared less than a million bucks.
By Macro Man's reckoning, that makes him either the worst evil genius or the biggest patsy we've seen in quite some time.
This is not, of course, to condone the sort of behaviour in which Mr. Sarao was allegedly engaging. Spamming phony orders is a reprehensible practice, and as a punter there is nothing more frustrating than trying to pay an offer that you see on screen only for it to magically disappear the moment you click "Buy."
However, given the prevalence of similar and even more morally dubious practices in single-stock world, it is bemusing to say the least that the only entity to be arrested (to Macro Man's knowledge, at least) is a small-fry punter who conveniently happens to be a foreign national. No mention at all has been made of the HFT algos, for whom spamming phony orders and front-running is a (very profitable) way of life.
When it comes to the Matrix, it appears that the regulators have thus far decided to take the blue pill.
Macro Man is hardly alone in his skepticism. He has unsurprisingly been chatting offline with a few mates with many years of experience at the coal face of financial risk taking, and no one has thought the situatio0n anything short of absurd. In the context of one of these conversations, one friend offered a suggestion on how to replace the badly-broken market mechanism with a few simple rules to replace many of the current Byzantine regulations:
1) If you are acting as an agent and an order for a client gets filled (full or partial), then the client gets filled at the execution price- no front-running allowed.
2) You cannot submit any orders for more size than your margin currently supports.
3) Any order submitted to an exchange or market must remain valid for a non-quantum length of time (call it one second) before it may be cancelled.
4) If you buy or sell something, you must make or take delivery within the specified amount of time. Failure to do so should incur a hefty financial penalty (scalable for repeat offenders) and a ban from further trading until delivery is made/taken.
To this, Macro Man added a fifth:
5) All participants trading on an exchange should incur the same exchange fees. Exceptions can be made for locals trading in open outcry pits.
Macro Man is curious what readers make of these suggestions and whether they have a few good ones of their own....
71 comments
Click here for comments@MM - Good post & I generally agree with you. Just to further clarify what you're saying, we have to be very careful in how we categorize HFT. A lot of legitimate trading and market-making is HFT, the problem is 'predatory HFT' - algos from HFT firms that front-run, spoof, cause stop runs via momentum ignition etc. The other problem lies with exchanges pandering to toxic HFT by having hidden order types, low fees for HFT and the like.
ReplySo, yes we need a solution. But a mandatory 1 sec delay is not the way forward IMO (it's too slow - even for manual traders). Some platforms utilise randomisation (e.g. ParFX) - that's ok, but also has some issues. I personally think limit orders should incur a very, very small fee - that would kill predatory HFT by wiping out all it's margins, in fact making it profit -ve. So say I pay $5/mio for a spot FX round-trip transaction, perhaps I could pay 1% of that (eg 5c) to open & close a limit order (even if it never got filled). That way I would likely only put in orders that I intend to trade on.
What really needs to happen is that exchange order types and fees need re-visiting. US regulators also need to get their act together. Good governance is THE cornerstone of financial markets in the developed world and it's rapidly becoming a joke.
Good points
ReplyMaximum order size should be measured and applied for each and every client. Some day trading guy should have limitations. I think the bank or corporation, member of the exchange, have a responsibility for their orders.
and just to be clear. I don't blame Mr Sarao.
ReplyAgreed MM, this is a complete farce and one of the rare occasions I find myself agreeing with ZH too :)
ReplyHowever, I would say that spoofing is less bad than front-running. In fact, to the extent that it keeps front-running in check, it's actually useful on the margin. Similar to having spiders in your garden. You may not like them per se, but their net contribution is positive, as they kill other bugs and pests.
This guy makes a similar argument:
http://www.bloombergview.com/articles/2015-01-23/high-frequency-trading-spoofers-and-front-running
@ theta
ReplySpoofing needs to incur the real risk of getting paid/given on one's orders, which is the point of maintaining a minimum order length.
Spoof orders that can axiomatically not be transacted upon have no business in a properly functioning market.
How about only best bid or offer are published. no size and no visibility of orders behind. Take away the information and you take away the value of it or ability to be spoofed by it
ReplyMM,
ReplyBut it does, someone could place a buy order for 5000 lots at 3 points above current level (or a market order for 5000 lots), which would instantly lift all his orders. It's unlikely, but theoretically possible.
Otherwise, if the argument is that these orders are impossible to trade, then they have no informational value anyway, in which case there's no spoofing! And in any case, the point remains that the only ones that are "spoofed" are the frontrunners, so to me it's a case of the lesser of 2 evils. You correctly said that it's very annoying to click "buy" and have the offer disappear. Well, I would add that it's also (if not more) annoying to have an order book of say 61 at 63, to place an order to buy at 62 and the next millisecond the screen is 64 at 66.
@theta
ReplyAs long as there is a real risk of getting lifted, then I am OK. Order that automatically cancel if other orders further up the queue get filled are clearly not intended to be filled.
As for having no informational content, well, when orders are marked 'Ha ha spoof! Ignore me!' , then you are right, they have no informational content.
If, however, u see say 3000 lots on offer over the next 2 ticks and so place a market order to buy 1500, and 2900 of those lots on offer are spoofs, your fill can be quite a bit worse than expected. Shame on you perhaps for not using a limit order...but I'd argue that in a well functioning market you should be able to pay an offer that you see with a market order!
(How'd you feel if you bought a litre of milk at the supermarket, and instead of $1.05 that the price tage says, you were charged $1.15 at the till? "I'm sorry sir, that price was a spoof, the real offer for you size was up here.")
Finally, note my proviso that phony orders have no business in a 'properly functioning market.' I think the current Matrix of algo front-running is pretty far from that.
I think we agree on things for the most part, but these little details are interesting to discuss (for me at least!)
on a (slight) tangent, why was a guy who made $50 MM trading in the last 4 yrs holed up with his parents in hounslow??
Replyi knew london was expensive, but holy..
MM I agree on both that we largely agree on the big picture, and that it's interesting to discuss the details.
ReplyWhen you place a market order are you sure it doesn't instantly take out all the existing order book for its size without giving them a chance to cancel? You may be right, I never place market orders even if I want to trade through levels I do limit at 2 or 3 ticks above, but I thought market orders would work the same as limit at infinity for example, i.e. take out instantly what's available.
@washed:
Reply"The risk manager calls him up and tells him, you know, 'Nav, what's going on? This position, you want to get out of it?' He was like, 'Well, what's the status on it?' 'Well, up 13 million dollars' and he was like, 'Ok just get me out.'
"Now a normal person would be jumping for joy, going ecstatic. But that's part of his job. He just came in to do that. That's what he did. His goal was to take on the trade that made sense, not because he wanted the money, not because it was greed."
Read more: http://www.businessinsider.com/traders-says-nav-sarao-had-balls-2015-4#ixzz3Y8sxrmdX
Kill the volume rebates. All of MM strategies are making a killing off of razor thin margins due to front-running and exchange rebates that everyone else pays for.
ReplyThe other major issue is that the exchanges are back-stopping the MMs (HFT shops) when their algos slip a disk. HFT shops should not be immune from the tail-risk they assume trading for razor-thin marings. Let them fail when they fail and the market will normalize again.
In other words, stop busting trades! Selective trade-busting is about as egregious an offence as you can find in this field.
Replyhttp://brontecapital.blogspot.in/2015/04/navinder-singh-sarao-our-spoofing-hero.html
Reply@ theta My understanding (and I am prepared to be corrected here) is that the spoofing algos are set up to automatically cancel if orders closer to the market get filled, and to re-establish a "safe" distance from the market. I suppose as you say, it is theoretically possible for these orders to get filled if someone places a limit order sufficiently far beyond the spoof order level that everything is done at the same time, but even then I think the closer orders get filled first, possibly giving the spoof orders time to get pulled.
ReplyThis is why my friend and I think mandating a minimum time limit for orders is a good idea. A nominal fee for placing orders would work well in single stock land, where there is a flurry of activity on quantum time scales and the value of each contract (share) is relatively small; given the leverage available in most financial futures, I am not sure how well it would work.
@ Dan the liquidity rebate nonsense is why I added rule #5 in the post; if you, I, or Johnny pension fund pay a fee for trading, so too should Jimmy HFT Parasite. Under no circumstances should a portion of our exchange fees be given to Jimmy so that he can trade in front of us.
(As an aside, has anyone else noticed how 'providing liquidity' has replaced 'making markets' in much of the debate? Making markets implies making two risk prices, whereas providing liquidity can simply represent spamming limit orders for the purpose of front-running legitimate flow.)
@MM 4:07 - You have hit the nail on the head here wrt the "maker-taker" model, rebates & phony liquidity. (If anyone is interested, Scott Paterson outlines these issues in a v readable format in his books "The Quants" & "Dark Pools").
ReplyThe thing about min-cancel times, intraday auctions etc is that while different they are also inherently gameable. If the goal is to prevent spam in the order books the best way to do this (IMHO of course) is through very small fees for all order operations. This would have to be a regulatory fee, not an exchange fee, to circumvent the conflicts of interest at the exchange level. I would also look at simplifying the exchange architecture, with removal of reg-nms.
ReplyBut honestly in the end I don't really think there is a problem. Markets are unstable with or without computers. We are starting from this invalid idea that the flash crash was a mistake - that tail risk does not regularly bite people in the ass, that market returns are somehow by law supposed to follow some normal distribution.
@ Mr. T.
ReplyExactly! The flash-crash should be a statistically rare event for HFT participants, but an inevitable event(s) no less. The market will clean up the HFT shops, if universal market rules are enforced. Since May 2010 there have been several occasions when the exchanges have busted trades to save HFT shops. AAPL in August 2013 is another one that comes to mind.
I think you can take it as read that "fat finger" or "obviously bad" trades will continue to get broken, giving the algos something of a get out of jail free card. As such, I prefer to cut to the chase and think up ways to eliminate parasitic trading altogether.
ReplyAs for Hempton's post, I don;t buy it. If someone breaks into your house and steals your TV, it's understandable that you're mad. But if someone then breaks into the thief's house and steals the same TV, while you may feel the cold satisfaction of schadenfreude it doesn't mean that you're getting your TV back. The second thief is not a hero.
MM, I think his point is that because he steals the thief's TV, the thief is going out of business, so you will get to keep your next TV.
ReplyI dare say that my example with the spider is better! :)
Still can't think of an actual solution to the problem though. I'm not sure a regulatory fee is the solution as it will act as an extra tax for the industry and especially legitimate market makers (imagine an options market maker quoting all strikes and adjusting prices with spot move all the time). Banning rebates is practically impossible (can you really dictate terms of business between private individuals, and if you can, should you?). Minimum time on the book sounds good, perhaps something like 100ms, too much for HFT, not really a problem for manual orders. But perhaps the best solution is not more regulation, but less. Let the algos fight each other with the spoofers and the front-runners engaging in a zero-sum game that doesn't give one of the two too much of an advantage (if you eliminate one, the other is free to steal other people's lunch, with both free to play, there's less damage for real traders/investors). And allow banks to carry inventory and make markets as opposed to putting more hurdles to them for fear of another 2008 (thus increasing the risk of another 1987). This will make algos less relevant, which will stabilise markets more than imposing all kinds of rules. Just a thought.
If you want to get rid of spoofing, like others have suggested, charge an order fee. Though for futures trading this wont totally solve the problem unless all exchanges agree, as volume can potentially migrate overseas.
ReplyAs someone who traded futures for a while, I dont think spoofing is really such a problem. You get used to it. Its more annoying than anything. Sure some orders get picked off but we are talking about ticks here. If you are worried about ticks, you are trading, and caveat emptor. For the investor (who the well functioning of markets is paramount for) a tick is meaningless unless you are in Eurodollar spreads like MM, but a tick in eMini's for real money is nothing.
The real problems IMO are in the single stock world (dark pools, quote stuffing, SMART orders), OTC bond markets, & Options markets (TD Ameritrade gets millions for routing their customer orders)
Spoofing in the S&P, Bunds or TNotes, really not a problem overall. And like Mr T said, these markets are volatile and are meant to have these big moves every once in a while to shake things up.
While leaving the market to be unfettered might sound good in principle, I am not sure how good a Wild West style marketplace (Gunfight at the Quantum Corral?) would be in practice. Actually, I am pretty sure that there would be more egregious abuses than we currently observe.
ReplyI do agree that banks should be allowed to act as principals, and would have absolutely no problem exempting market makers who commit to making 2 way risk prices from any exchange/regulatory fees.
I do strongly feel that customer/price taking business should be pari passu when it comes to exchange fees, etc., insofar as exchanges are in many ways public utilities (and indeed regulated as such.) This includes front-running 'liquidity providers' and the HFTs that arb discrepancies between prices on Tom's exchange versus those on Dick's and Harry's. (In stock space at least, a lot of the nonsense could be eliminated by dispensing with the plethora of exchanges and dark pools, etc...but that seems unlikely to happen.)
@ Abee, I'd agree that the most egregious abuses are in single stock land (which makes the scapegoating of a futures guy who made something like a quarter-point per contract he traded on Flash Crash day all the more absurd.)
ReplyThat having been said, it would be nice to not see attempts to trade on the bid/offer fail, or to be routinely ticked by Agent Smith when leaving a mid market order in futures as well.
To use theta's analogy...just because the spiders in your kitchen kill the ants and flies doesn't mean you want to get bitten by one! For choice, you'd rather have no critters at all.
http://www.zerohedge.com/news/2015-04-23/why-sarao-flash-crash-patsy-he-threatened-expose-mass-manipulation-high-frequency-ne
ReplyAgreed that no critters at all is best.
ReplyOn a completely unrelated note, Mrs theta just saw me reading the blog and she pointed out that the name Macro Man reminded her of the following song:
https://www.youtube.com/watch?v=AO43p2Wqc08
@ Anon, I really don't believe that Sarao is an 'old-fashioned, point and click trader' at all...in addition to the evidence of the existence of his spoofing algo, I don;t believe that anyone can consistently make money playing for a quarter tick profit as a manual entry proposition.
ReplyI do not of one very high volume shop in Chicago that has traders manually clicking for *maximum* +/-1pt ES rides as well as other products/markets/exchanges. Win or lose, they are out at 1pt max. They make it work because of their exchange rebates.
Reply*know
Replytwo exchanges then. one an electronic snipefest and the other an well regulated orderly more old fashioned friendly slow but trustable centre of good manners. Then let mkt forces decide which users turn to. Fiver says its the tighter prices one, snipefest or not.
ReplyI'm an independent algo trader, self-financed, a few of my strategies are what would be considered "hft". The primary one: market making.
ReplyI actually agree with all points with the exception of number 3, and on the fence on 5. Wrt 3, the desire can be broken down to essentially: "market makers must stay in one place long enough for me to smash them." An obvious example, what market maker in his right mind is going to tightly quote SPY in forced seconds with so many derivatives quoting with non forced seconds? Price discovery will move solely to futures, SPY will no longer have any market share, and anyone quoting there will have spreads so wide as to not be useful. This example can be expanded nearly infinitely as most products are inter-related in some form, and most market making books make use of those inter-relations to keep their deltas low. So, are you going to force 1 second quotes for every product in existence?
As to five, I'm not sure I'm seeing the motivation for this. You want the exchanges to try and NOT attract liquidity for individuals/firms who want to provide it? If this were x'd, spreads would simply widen by that amount. I don't think most of you realize how thin short term trading profits are... RAZOR.
Love the blog
@ Algo Re point 3- yeah, I think the intention would be for anyone leaving an order (as opposed to making a 2 way price) in any product to have to have it remain valid for a non-quantum amount of time (second, half a second, I dunno) to discourage spamming/spoofing etc.
ReplyAs for point 5- as I noted in the comments, I would be happy for market makers who commit to making 2 way risk prices to be exempt from the fees. 'Customers', be they macro men or quantum parasites, who trade via orders, should be pari passu to discourage spamming.
Might this result in some wider prices? Yes, though on the time scale of 95% of investors (by number, not volume) it won't make a discernible difference.
If there is the ancillary benefit of a more robust, trustworthy market microstructure that is not at the whim of occasionally uncontrollable robots, it's a price I would happily pay.
http://www.bloombergview.com/articles/2015-04-24/michael-lewis-has-questions-about-flash-crash
ReplyThe main problem with a number of the proposed changes is that while they do attempt to tackle clearly defined problems (e.g. some mkt participants having a speed advantage), they do introduce an element of whack-a-mole to the game, and simply expose a different flank for smart people to plunder some edge from.
ReplyFor example, the minimum order lifespan is a very worthwhile attempt to get everyone to pari passu wrt speed of access. This is a great idea, as the effect of order instruction latency is very much non-linear and really (really) favours the guy colocated at the exchange - he can't strictly speaking "front run" you (this term is misused here, I think), but he can react to the same market before you can. However, what you now have is the one product in the world that is correlated to many others, but is by definition slower! As a market maker, I'll have no interest in market making it with a tight spread, but would rather smash it to spread the risk off products on other exchanges... as you can see, this will soon lead to a wider spread on our original product. So perhaps it's not quite the right way to level the playing field wrt speed. The one way that forces more "thought" into the process of order placement is perhaps limiting the number of order instructions that can be sent per second, although that still leaves the colocated guy, or the guy who can afford to pay for more than one user, at an advantage...you see what I mean about mole-whacking?
The per-order fee is a nonsensical proposition. Do you want to have any market makers/locals at all? If you wisely do, you can't have per-order fees, however small, as there's a degree of automation in all meaningfully large shops and a per-order fee is again a quick route to a wider spread. It's a theme: 0-tick spreads have a price as someone carries that risk for you. They need to get paid for that risk.
The one thing that I was very surprised to not see mentioned was the nature of the matching engine. As short sterling has gloriously shown, if you favour the big boys (pro rata), there's soon just one big boy remaining. Just as selective busting of trades and prohibitive cost of colocation, this to me is probably the most unjust rig that is currently in existence. First-come-first-served regardless of size (aka FIFO) is such a no-brainer to levelling the playing field to me. I'd like to hear some arguments about that if we can.
The trouble is that it's hard to tell an old-style local from Evil HFT LLC these days, as they are all electronic. All you can do is set the conditions. This needs to be done, but has to be done extremely carefully as sweeping changes will just expose different edges that would be exploited as quickly as the old ones by just the same people.
Info on Predatory HFT:
Replyhttp://www.bettermarkets.com/blogs/%E2%80%9Cflash-boys%E2%80%9D-puts-flashlight-dark-predatory-hft-trading#.VTonGSGrTsE
It's clear that predatory HFT is actually reducing real liquidity and thus is the real cause of the various flash crashes.
ReplyBanks, pension funds, hedge funds & other asset managers should be taking legal action against these HFT firms (and possible the exchanges) for the lost revenues they have incurred. Regulators & government should also look to bring criminal charges against these HFT firms.
Predatory HFT firms make billions of dollars a year, with almost no losing days (a sign that their activities are not legitimate).
Unhappy about the extradition of Sarao? Then sign this petition:
Replyhttps://www.change.org/p/david-cameron-mp-to-not-extradite-navinder-singh-sarao-to-the-us
The per-order fee is a nonsensical proposition. Do you want to have any market makers/locals at all? If you wisely do, you can't have per-order fees, however small, as there's a degree of automation in all meaningfully large shops and a per-order fee is again a quick route to a wider spread. It's a theme: 0-tick spreads have a price as someone carries that risk for you. They need to get paid for that risk.
ReplyI think people are incorrectly equating spreads with transaction costs. Costs as measured using full-order standards such as implementation shortfall are (loosely) negatively correlated with spreads. Yes, thats right, as spreads get tighter our total costs go up. How is that possible? Because the way "market makers" are paid for that risk is through information leakage that allows them to predict autocorrelated intraday orders as part of say a full-day vwap, and you get crushed on the opportunity cost side of the IS equation.
Consider your typical scenario - you submit a full-day vwap order that gets chunked up. Each chunk first pings a dark pool to try for midpoint pricing. It gets partially filled at midpoint (yay!) for 3 shares, but then mysteriously the offers drop on the lit markets, and the remainder of the bin either goes unexecuted or gets hit with some adverse selection. Repeat all day.
You could say "your algo is bad, dont route to dark pools". So say you see 1000 on the ask in a lit market, you try to get filled and the same thing happens, through the magic of reg-nms.
Given the current ratios of orders-to-trades its obvious at least to this market participant that the vast majority of the orders are there not for the purpose of trading but for detecting flow. If we are buying 1k and the market is 1c spread and there is 100 on the ask its unlikely the algo will cross the spread for a partial fill. If its showing 1k on the ask, it may very well cross the spread. So it reaches for the 1,000, but only gets 100 before the rest are cancelled. How is that different that Nav's spoofing?
And all this happens in the context of 1c spreads.
I may be missing something, but I think the per-order fee would solve much of this. It may result in higher spreads, but that does not mean higher costs.
Looks like the really did it...
Reply"The European Central Bank started buying covered bonds with negative yields"
Hey MM, could I temporarily distract you from your project of toppling Michael Lewis as flash trading commentator extraordinaire, and get your opinion on EURUSD? We seem to be getting into a weird twilight zone for data where we expected sequentially weak US and sequentially stronger global macro, and instead seem to have encountered weak everywhere macro - interested in your thoughts.
ReplySorry to not contribute to the other thread - I am still adjusting to my pin hole camera, so the point and click discussion is a bit haaafalootin like they say down here in good ol' texas.
SDRT to apply to all orders transmitted to an exchange or dark pool, no market maker exemption.
ReplyThat would level the playing field and capture some of the upside from their risk pollution to society at large.
+1 Mr. T
ReplyThe ultimate value of "providing liquidity" to these quant shops is that they are sampling real liquidity that informs their proprietary trading (front-running).
Pull the rebates and the front-running will evaporate! Twenty years ago, the pits were full of jocks who could front-run based upon the sentiment in the pit and their elbowing/shouting capability.
re: order-routing, that's really a broker issue.
http://www.bloomberg.com/news/articles/2015-04-23/trader-tied-to-flash-crash-says-he-simply-changed-his-mind-a-lot
ReplyOk now it is confirmed, Nav Sarao, is no HFT. First he use Trading Technologies (TT) which while it is the the industry standard, it is not even in the realm of HFT. Those guys almost always build their own and co-locate servers at the exchanges. TT is for pro's but not the HFT computer geeks.
As to the question can someone make a living looking for ticks, it is very possible. Not as easy as it used to be but there are LOTS of ppl doing it. I used to be one of them.
make 5 ticks a day on a 1000 in the eminis' there is your 60K a day. Not saying its easy and trading a 1000 clip is probably near the top of what you could do (now its lower but in 2006-2011 you could).
You dont get any market making fees in Futures. You get commission rebates, yes, but you still pay commissions. And Nav probably paid a lot in commissions if i guess at his trading style.
If he was using TT, his spoofing algo was probably very unsophisticated and not what most ppl would consider HFT. I guess spoofing is still illegal, but it is so widespread to jail this one bloke seems unfair. Make it clear to the whole market.
There are so many other cooked things in the market (ppl playing ahead of news, earnings, stop hunting etc) that going after spoofing in the eMini's, give me a break.
I want to know anyone who has trade the emini's with an order size under 5,000 lots who has seen more than 1 point (4 ticks) slippage. It doesnt happen.
Nothing has fundamentally changed it's just that the signifiers can't be measured with the human ear, the trades aren't phoned in and written on cards with ample time for people to talk about it all in between.
Reply@ washedup
ReplyI am still small short, and per a post last week added some short cable on Wednesday when it popped over 1.50 (out of court for the time being, of course.)
I remain of the view that the spring will see a revival in the US data, but the longer that that persists without materializing the more tenuous the conviction must be. That being said, it's still too early to get a proper flavour for what April has been like.
The problem with euro in the short run is that people are still short and there remains Greek headline risk. I personally think any 'deal' will be viewed in much the same was as the last one a couple of months ago- a temporary band-aid over a gaping mortal wound.
So while there is certainly risk of a gap higher if/when a deal is announced, I suspect it will be a fade.
Generally, however, I find the market is generally treading water across many assets these days, either searching for a new theme or waiting for muddied waters to clear. If and as clarity begins to emerge both myself and others will feel more comfortable deploying more risk, but for the time being I am trying to keep my position sizes very very social.
much appreciated MM - thx
Reply@washed,
ReplyYou understand Punjabi?
https://www.youtube.com/watch?v=SWI8dHWK6oA
I do understand a little punjabi, and if ur telling me this love ballad crooner in the video is the same bloke who made getco and jump trading pee in their pants (or not, wtf do i know abt such stuff), I am looking at him in a whole new light.
ReplyI seriously wouldn't worry bat this guys future - best case he gets hired by SAC to catch them up to the new ways of making money - worst case he does a jordon belfort and returns in a catch me if u can type stint at the CFTC.
@washed,
ReplyI am with MM -
It would be nice not to have 'bugs' in the system!
The video..just fun stuff!
http://news.forexlive.com/!/fed-flouts-congressional-request-for-leakers-20150424
ReplyHave not heard much from Bullard about the need for more QE lately - a bit surprising considering the weak economic data? The fed sure is quiet when the markets are rallying.
ReplyI would weigh in on the great short-bund debate with a "no thanks" vote. If I've learned anything from the last 7 years its don't fight the fed. If Draghi wants the entire german curve @ -20, thats what he will get. Also, I suspect we will see a further drop in the deposit rate, or a change in stance about what is eligible for purchase. So much credibility has been invested in ECB completing their program, I suspect even if the market front-ran them to the point of -100 they would continue to do "whatever it takes". Lastly, given all the talk of real shortages, that could make for a nasty surprise one morning when short a pile of GBL's.
A 130ms delay, the time for light to circumnavigate the earth, would put everyone on a more equal footing more or less. At least until a hedge fund tunnels through the earth to shorten the path to the exchange. Current tight spread's that are illusionary probably mean that risk is currently being underpriced causing occasional and almost instantaneous spikes in volatility which are conveniently ironed out by breaking trades when required.
ReplyAnon @ 10:37:
Reply130ms delay from what point in time? And delay of what? Arrival of order instruction at the exchange? The guy who is colocated is still going to get to submit his order before you. It's just that both of you will now have 130ms to make a cup of tea and come back to watch him get filled first.
This would also delay your product compared to all other correlated products traded elsewhere, but actually, having thought about it, I don't think that'd cause anything more than extreme confusion, as the artificial delay would just make it look like the market is still quoted at a price when the orders are already in the process of being cancelled (e.g. just after economic data events).
Randomising the delay between 0-130ms is also a bad idea as it'll just shift the play into a more statistical land with an average delay, leaving the above two problems in place.
The point I'm trying to make is that tinkering with complex systems exposes side effects you haven't thought of, that will bite you in the arse. This is as true of exchanges as it is of monetary policy. Keep things as simple as possible and create as equal access as you can (e.g FIFO, flat fees, one point of access to the exchange for all, the same limit on number of order instructions per unit of time for all) and you stand a chance of a creating a more-or-less fair market. You will never achieve 100% fairness. Mainly because fairness is not quantifiable ;)
Too big to fail!
ReplyToo big to jail!
"Gen. Petraeus: Too Big To Jail, Too Hypocritical To Shame".
http://davidstockmanscontracorner.com/gen-petraeus-too-big-to-jail-too-hypocritical-to-shame/
Re: EURUSD, it's all about relativity here.
ReplyWill a couple of randomly occurring soft patches in US be enough to outweigh chronically negative yields in 90% of the EZ (and without forgetting that Draghi might actually be willing to trample yield curves further underground)? And would a Greek kick-the-can deal be an ancillary contributor along with the soft patches in US? In the short term yes perhaps, especially should they conjunction together sequentially giving appearance of one side getting stronger and the other weaker, but it shouldn't be long lasting.
The hill is getting progressively steeper and it would seem that with every kick the can takes, it's able to roll a shorter distance under it's own momentum. What is really needed is a long term debt restructuring and I just don't see how that could happen. Even if it somehow miraculously did happen, it would probably be a rocket ship with very limited fuel for the euro as people realize that the kind of EM driven export growth the EZ enjoyed for the first decade of its existance is never going to come back, and which was a very important prerequisite to be able to sweep all the design flaws under the mattress for a long period of time.
Now fast forward to present, the growth disappeared and here we are stuck with all the original flaws and the current attempt to hold it together is to crush the bond yields, because it's the only thing that has a chance of keeping the periphery content enough so they can continue to fund all happy clappy oversized social-free-for-all budgets and public servant benefits. But no matter how much makeup is applied on top of it, it's all fundamentally a screwed set-up. Atleast the US is solid and unified (enough) that it won't break up with the first real bump down the road without the heavy use of smoke and mirrors.
I'll jump in MM's train here, heavy short positions and short term headline risk on the can kicking, but whatever it is going to be it's going to fade shortly. If DXY continues to fall, which is probably a big if in anything other than the very short time frame, it should more or less allow oil, gold and other risk stuff to climb.
"The Biggest EURUSD Resistance Test of the Year"
Replyhttp://www.dailyfx.com/forex/technical/article/fx_technical_weekly/2015/04/24/The-Biggest-EURUSD-Resistance-Test-of-the-Year.html
Neither disappointing earnings nor econ data have derailed this market for years. And only taper tantrum has FED been viewed as hindrance.
ReplyAll very happy-clappy this morning at first. VIX at 12.5, the most important earnings report in the galaxy after the close and a Fed meeting ahead. What could possibly go wrong?
ReplyBucky Breakdown, for one thing. DX dropped below its 50 day average at 97, for the first time since last July (when it was at 80). We have warned in this space of the effects of an unwind in the Long Dollar trade and its associated leveraged derivatives...
hey left - I though you were silent because like me, you are clueless (and therefore speechless) in the arcane field of high frequency trading and associated dark arts - good to know you were actually off poring over dollar charts.
ReplyWelcome back to HFTman.blogspot.com - clearly I exited too early on commodities/EMs - nice outperformance from that gang today like u said.
And biotechs - ouch - atleast the traders in shanghai (I'm told) lack high school diplomas - whats the excuse for wharton MBAs who currently hold that basket at 500X lack of earnings??
EM FX has settled in, as the Euro stopped, Oil bounced and SHCOMP rallied, but on the ground not much has changed IMO
ReplyOil I think will be capped at 75 so I am not a big buyer here as equities have already priced it in. SHCOMP is all about pumping up the index so banks can raise equity. Pure casino, i cashed out. Probably it will go higher before it comes down, who knows but its not going higher bc the economy is getting that much better. Extrapolating past china stimilus to this one I think are misplaced, commodities are not going to see a big increase in demand. so i am still warry here. everything seem a little complacent but I still think sell in may is the way to go for SPX
Finally...people are starting to realize how sick this man is...
Reply"Greece’s outspoken finance minister Yanis Varoufakis has been sidelined after three months of fruitless talks with international creditors to unlock €7.2bn in bailout funds, heartening investors and sparking a rally on the Athens stock market."
Clearly, he had a breakdown. Just watching him give a speech or respond during an interview should give someone pause...I believe he suffers from borderline personality disorder.
Yanis Varoufakis @yanisvaroufakis
ReplyFDR, 1936: "They are unanimous in their hate for me; and I welcome their hatred." A quotation close to my heart (& reality) these days
Good work Abe....are you corresponding with Yanis?
ReplyJapanese YoY Retail Sales:
Previous: -1.7%
Forecast: -7.3%
Actual: -9.7%
MoM Retail Sales:
February: +0.7%
Forecast: +0.6%
March: -1.9%
Congratulations to Baofeng, 25 consecutive days of limit up! IPO @ 7.14 on Mar 20, now at 101.27
Replyhttp://tinyurl.com/npmw2fn
P.S. This is completely normal market behavior.
Yanis-V is no Theseus and has never been someone who could close the deal. When has an academic ever had to close a deal in the wild? Reading and writing about game theory does not mean all of the other players at the table don't eat, sleep and breath wild negotiations. Even Ben Shalom needed Hank to close the deal.
ReplyIn the end, Yanis-V will probably be torn apart by his own people much like Orpheus. He should have known better than to sign up for something he wasn't capable of finishing.
Italy debt management chief {MARIA CANNATA} defends use of derivatives which have cost >€12bn over the past 4years
ReplyDon't you just get sick reading this crap?
It's soooooo tough managing a 2.17 trillion euro debt pile of paper.
Gee, MM when does this end?
Agree on oil cap @ 75 give or take a little. That cap should be justified by the fact that even though shale oil was in trouble a while ago with their over sized debt burden, the technology itself should be proven and quite profitable at those levels and would result in any (major portion) possible shutdown supply becoming online again and then we're back on course heading to square #1. Of course lets not forget it might overshoot again because it's a commodity.
ReplySo if oil is one of the major factors in inflation expectations and when it reaches close to that level, it might be a good idea to again sell the anti-dollar vehicles then and get back in TLT and bucky again as the "transitory effects dissipates"/inflation expectations reach their peak again, on a wrongly concluded assumption. For now just sit on risk and wait while DXY makes its little mandatory spin downwards, depending on Fed it might accelerate a bit but near 75$ should still be a good guideline to put on reverse.
DXY is down again (looks to the Heavens for thunderbolts). I know that's unpopular opinion. At what point does the reflexivity set in to reverse the reflexivity that got us up here in the first place? A break below 95?
ReplyMM will be on in a minute to smugly do the "stopped clock" joke ("even a stopped clock/LB is right twice a day). Actually I'd be happy to be right twice a day, or even twice a year on big issues.
Any chart wizards want to call this recent DX retracement? Chart support at DX 95, anyone? EURUSD 1.12 has been bandied about.
Piss taking apart, our EM and European longs are already looking quite tasty...
DXY was certainly crowded, but is a lot less crowded than before - at some point they will over do it the other way, and with an oversold RSI I don't think we are far from there (say 95.5-96) - there has been a pretty decent pickup in inflation breakevens last few days, so the case for the fed to remain dovish with equities at record highs and crude materially off the lows is looking iffier by the day.
ReplyAs long as u don't forget the EM/european rally is a flow related, counter-trend move, I suppose u are fine.
http://www.telegraph.co.uk/finance/comment/jeremy-warner/11569329/Jeremy-Warner-Negative-interest-rates-put-world-on-course-for-biggest-mass-default-in-history.html
ReplySome extracts:
The flip side of the cheap money story is soaring asset prices...Eventually, there will be a massive correction, in which creditors will suffer sickening losses.
We live in an “extend and pretend” world in which economies pathetically fight between themselves for any scraps of demand. One burst of money printing is met by another in an ultimately futile, zero-sum game of competitive currency devaluation...Both Keynsian and monetary economics seem to be in some kind of end game. What comes next is anyone’s guess.
Johnny Retail here... almost every day lately one of my U.S. holdings goes up 10+%. MSFT, RCII, etc, etc.
ReplyA bit frothy, innit?
Johnny Retail #2 here, has had my ass handed to me twice in the last 15 years and can't stomach getting back in "full flight". Looks frothy, smells frothy, but also makes me feel like an idiot every day.....
Reply- Whammer