gd1

Does that not lead you to question your assumption that the war was "for oil in Iraq"?

Huh? The rule is you can't place orders that you have no intention of executing at the time you place them. You're perfectly entitled to change your mind afterwards, or adjust your price as new information becomes available. You just can't place orders when it is your goal to not have them execute, and that was your goal before you even sent them.

"For those playing along, here's the metaphor. Joe goes to market to buy sheep. Bill knows Joe is going so he sends a fast runner ahead of him to buy the... "

Seriously, how many times have we discussed this issue on this site and we still get this bullshit. Bill doesn't know Joe is going. He doesn't. Get it through your thick heads.

Academia really is a massive wank

>JIT'ed java code will run at native C performance... because it essentially is native C once it's been JIT'ed

No way. It may be native C, but it's not native C you've hand-coded yourself.

I've got an idea. CAPTCHAs could be replaced with suitable items of work from Amazon's Mechanical Turk. It stands to reason that these can't be solved by current computer vision algos, or else they wouldn't be on Mechanical Turk in the first place(?), and the side benefit is that the site owner can collect the 5 cents or whatever the going rate is.

Can't say I agree with that. Soccer has a state too, just because they don't pause the game after every completed pass for a few adverts, a timeout, and 12 replays, doesn't mean it doesn't have state.

Strange argument. How does modern soccer (quite attacking these days) compare to the number of goals per game in ice hockey, touchdowns per game in the NFL, or home runs in a baseball game?

So you've accepted that greater speeds allow market makers to quote tighter, so the answer is pretty simple. The tighter a market maker is quoting the less money he is making per transaction (smaller spread), so less money is leaving the system and going to middlemen. The speed allows them to undercut other market-makers and steal their 'flow' (customers) while still being able to avoid predators, and so they offer the service cheaper. The service being the provision of liquidity to bridge time gaps in a continuous time market.

If you assume that the number of 'real' trader (not middlemen) who want to buy or sell is constant (not necessarily true, but assume it for a moment), then it stands to reason that the tighter the spread, the less the market makers are profiting. At a small enough spread, they make no profit, since the profit from the flow on both sides is cancelled out by the losses from adverse selection (trading when they couldn't cancel in time - being victim to a predator). But if they can increase their speed, and minimize their adverse selection losses, then they can quote tighter and still make a profit. And do it with machines, and cut overheads of hiring humans, and you can quote even tighter since you need less trading profit to make a net profit. So this is the value generated, automation + speed = the smallest amount of frictional costs being extracted from the market.

The speed for a market maker (largely passive and posting limit orders) isn't about how fast you can place a limit order, it is about how fast you can cancel it. It is a defensive mechanism against predators who are even quicker (arbitrageurs, aggressive HFTs).

Imagine that there is a "true price" of some security at price x, and your bid and your offer are positioned around that true price at x - d (your bid) and x + d (your offer). So your spread is 2d. Then some external (public) event happens, which moves the true price by an amount greater than d. Let's say it is a big event, and the true price falls by 4d. Now there is a race. If you don't manage to cancel your bid in time, a predator will 'pick you off', take out your bid, and they'll sell to you at (x - d) when the true price is now (x - 4d). The predator has made an immediate (paper) profit of 3d, and you've made a loss of the same amount. If you're fast enough, you can successfully cancel your bid and repost it further down the book at x - 5d.

Now, as a market maker, a liquidity provider, you could increase your safety margin by quoting a wider spread, setting your bid and offer at +- 2d instead of +- 1d. Now it takes a larger move before you're in danger of being "picked off", but you'll also attract less customers and make less profits. Speed (latency) is directly correlated to how tight a spread a market maker can quote.

>The 7-11 provides the trade functions "assortment of goods" and "transport" and as such is valuable.

"Large supermarket chain" also provides the trade functions "assortment of goods" and "transport", so what exactly are you paying your extra 50c for? Liquidity.

What a bunch of assholes those 7-11 types are. Charging us an extra 50c for milk. The nerve.

Again, back to my milk analogy further up, in theory you could get your milk direct from the farmer and cut out the middleman. I'm sure some of them go to markets and set up stalls and such. But the reality is, it isn't the farmer's game to do that. Every moment he holds that milk he is carrying inventory risk (it could spoil). And if a buyer doesn't turn up, then it definitely will spoil. Instead he offloads it to a middleman who is equipped to handle it (refrigerated tankers, etc.). Now in this real world example, the bulk of what the middleman is paid for is value-add (transport, pastuerization, bottling, etc.), but there is also the payment for liquidity. These companies (i.e. Fonterra) will be there to take your milk every morning. So some of what you pay for is that liquidity.

The people who need the liquidity are the people who willingly take the liquidity. Do you think someone puts a gun to their head and forces them to submit a market order?

It's a little bit like saying the guy who runs your 7-11 is a scumbag because he charges an extra 50c for a carton of milk, and "who needs that liquidity, they could just wait until <insert large supermarket chain> opens and get their milk there". Sure, but they might want the milk now. And no one forces them to get it from the 7-11. It's a free market.

An article written by a journalist, about a book written by a journalist. And neither of them have a clue. Layers upon layers of idiocy.

Didn't I just explain it? I can't do a descending sort of every store on my local high street based on their success so far. These things snowball on the internet because the sorting and quantifying process lends itself to viral snowballing.

Not to state the bleeding obvious, but isn't real life a little bit different to the internet? If I want to see which kickstarters have had a modicum of success, I just do a descending sort. It doesn't work like that in real life.

What are you on about? Limitless energy is rained down on us by the sun, and matter itself is conserved at the elemental level (barring nuclear processes, which are minimal on our planet). What do you suppose we are running out of?

You send your burst a few milliseconds before an event. It chokes up C (all of those depth updates sit in his queue and he's busy for the next 50 millis churning through them). Then both you and A are free to pick him off.

The thing is, you know to ignore the burst on the public feed, since you know it is coming (you created it). You still have to process the order book updates, but you can skip your decision logic. So any participant who has an internal latency (packet decode + orderbook update + decision logic) greater than 1 exchange cycle, and who doesn't have an "out" clause to stop making decisions if it is falling behind, will be vulnerable.

Like I said, I don't buy it, but it is possible in theory.

>And given that quote stuffing doesn't actually hurt anyone, it's not like stopping it is a high priority.

It does hurt people in some way. It increases the amount of data we need to handle. More storage space and more processing required.

And you've been a bit disingenuous in your article - there is a theory behind the "2.???" part. The theory is that slower participants are overwhelmed by the quote traffic and end up lagging, and are then taken advantage of. I don't personally believe it, and I'm sure 99% of these things are just feedback loops, but it is theoretically possible. The faster exchanges are cycling in say 20-30 micros these days, so any participant who can't match that rate will easily fall behind. Who knows, weaker participants may still be taking 1-2 millis to handle an update. So just before some known vol event, say an economic figure, you flood the exchange and lag the other participants by say 50 millis, then pick them off after the new information arrives.

Sorry, missed that bit

Obviously, well written HFTs avoid this by introducing some hysteresis to the signal. i.e. Insert an order when the signal crosses 0.5, but then only delete it if the signal falls below 0.4.

Chris: You've also missed the most common cause of these issues in my experience - feedback from identifying your own orders in the public stream. HFTs need to filter their own orders out of the orderbook they are trading on, or a feedback loop can begin. This can be difficult, since some exchanges don't provide a clear way to identify your own order, or it may be the case that you receive the public UDP packet before the private TCP packet that contains the ID required to match the two.

For example (a basic one), say I have an algo that calculates the midpoint based on the best bid and ask volumes. And can only insert within x distance of the midpoint. I add a bid to the top of the book for a 1 lot, get the UDP public depth update, don't know whether this is my order or someone elses. So I re-calculate the midpoint, which pushes the midpoint upwards (greater bid volume), which means I then pull my bid, which means I recalculate the mid again, then re-insert, etc. etc....

I just checked my database, 69235709557 rows. And I wouldn't even consider that "big data".

>and a beach house

Surely you mean a bach bro?

How the hell does anyone determine that? If someone walked out of Id during Quake III development with Carmacks's inverse square root (http://stackoverflow.com/questions/1349542/john-carmacks-unu...), would that be a trade secret? It's not like an inverse square root is a secret, but what about doing it faster than anyone else?? Writing the best 3D engine was part of their success, and writing the fastest order routing/exchange is a big part of HFT. He took source code from GS to go to a competitor that had offered him a lot of money, it's hard to have much sympathy for the guy.

>none of the code involved trading strategies (the really valuable stuff)

Wrong. In HFT, the 'strategies' are the least complex part. They are often dead simple. A well written (fast) exchange connection is at least 70% of the game.

He's a thief and he got caught, but the HN community don't like that narrative.

What do shares have to do with anything? They are a small part of the picture. The spread on the e-mini S&P is currently 0.25, and it is trading at 1800. So 1/7200 is roughly 1.5 bps. This will increase the spread by (at least) 6 bps to 7.5bps, then charge you another 3. So spreads five times bigger. Instead of trading say 1810.00/1810.25, it will be trading 1810.00/1811.25. Do you get it yet? No one will trade. And the comical part is that they project supposed revenues based on current trading volumes.

Also, I've missed the best part so far, when I mentioned that you get charged three times in my first post, that was just to get into a trade. So end-user will get charged at least 6 times per roundtrip (twice directly, four times indirectly through widened spreads).

"In fact, we can show that the spreads must widen to be larger than the tax itself being applied. If there’s a tax of 3 bps on a trade then no one’s going to trade at a spread of less than 3 bps."

6bps. To market make a spread you have to buy AND sell. Tax is charged twice at each end.

The intermediaries are charged the tax twice, which means the end consumer (investors, pension funds, whatever) will be charged (at least) three times. Once directly, and twice indirectly because market makers will have to widen their spreads by at least that amount.

Heh, true. I guess what I was saying is that if you freeze a basketball score at 20-18 in the second quarter, and at that point the odds are 52-48 in favour of team leading by 2, and then neither team scores for the rest of the match, then the odds will asymptotically approach 100-0 as you approach the final buzzer. Even though no one is scoring. By the variance measure used here, the second half will appear more exciting, even though nobody scores. Reminiscent of football (soccer) perhaps. The fact that first halfs have less variance is just evidence that they are further from conclusion, not necessarily that they are less exciting, but then maybe the two are linked.

Why did you edit away your claim that it is to the benefit of the customers? Being a market-maker and being able to void any trades that occur during volatile periods sounds like a dream to me. No need to worry about adverse selection eh? Just printing cash with fat bid-ask spreads and if anyone dares to get the jump on you, they get voided. I wish I could do that in the real markets.

How do these betting sites handle real-time events? I believe there was a case recently where a man at the Australian Open tennis in Melbourne was arrested for transmitting point information outside the stadium before it could be broadcast on television (there is always broadcast delay), which obviously could give you a big advantage. I think European football ones go into a vol auction (pause betting) when a goal is scored? But you could have a guy sitting in the stadium, wired up with a buzzer to press when a striker goes one on one with a goalie, or a penalty is awarded, and then just go all-in on the market? The whole thing is a can of worms.