You realize you are promoting a routine mutilating surgery for possible (but unproven) reduction in the likelihood of various unlikely diseases? Other countries in the world look at the US and laugh.
Admittedly, I haven't rtfa (this IS slashdot), but I'd suggest that there is likely a causative factor in the kinds of parents that choose to circumcise in those parts of the world. I can say for sure that I get much of my sexual stimulation from the foreskin itself. Denying that of someone because of possible issues later is ridiculous. It's also against the Hippocratic "do no harm" tenet.
Pediatrics is the only branch of medicine that routinely performs unnecessary surgeries. The US is almost alone in routinely performing circumcision for non-religious reasons. Canadian pediatrics recommends against it, IIRC. And every year circumcisions cause disfigurement and death.
You count to do multiplication?? Really? What about powers? How do you even know what to do to get to the result? The procedures must be memorized. The symbols. Shortcuts. If you've ever worked with kids you'll know that you can't merely teach a method and let them go. They get stuck at the first step. They have to learn facts along with the method.
How fast can you count the first billion natural numbers? Time yourself... I'm curious.
You do need to memorize some results. You don't, for example, do repeated increments in your head to add, do you?...and then repeat that process to do multiplication? Most people will look at 5*5 and recall 25 without going thru a tedious counting process. Ymmv Educators probably rely a little to much on rote learning, but a certain amount is crucial.
That sounds like a rhetorical question, but if you think an unvaccinated person cannot make vaccinated people sick, then you are seriously ignorant. Please go learn about vacination, the various vaccines in common usage, their efficacy, and the diseases they help prevent. This is way too serious an issue to go making this kind of comment without actually knowing anything about the subjevt matter.
It's true that some market makers pull or widen significantly before certain announcements. But energy swings so violently after news that they can't afford to stay in there. It's only for a few minutes. Without HFT the market makers and arbitrageurs would be there anyway. And yes, actually, even flesh and blood market makers would widen the spreads leading up to news - otherwise they would lose money. And I don't mean a little during the swing, I mean overall the losses would outway any profit. Market makers are just ad focused on avoiding losing money from big moves as they are focused on making a tiny profit on the spread. You are sure that HFT MUST be bad, but you don't understand it, and frankly, the evidence is against you.
All orders are attributed to a specific trader. Although an algo selected when to order, what price and so on, the trader takes responsibility for every order. And it actually is enforced. I have first hand knowledge of inquiries and punitive actions. Behaviors that are considered manipulative are punished very severely. They aren't good for the exchange, SRO or marketplace.
Fact: the biggest crashes over the past couple of years have occurred when HFT pulled out due to bad information and impossible quotes.
Question: Where did those impossible quotes come from? Why were they impossible? What prompted such quotes? Were the algo quotes, or human quotes? Was HFT any significant factor in why such impossible quotes were made?
The quotes I'm referring to were market data from the 2010 flash crash incident. That market data was published by NYSE. The prices were correct, but the data had been delayed. But the bad part was the timestamps on the data indicated they were fresh. It appears that NYSE queued up market data, but then timestamped it on exit from the queue. This created the appearance of huge arbitration opportunities between NYSE and the other equities venues. Those opportunities really didn't exist, but that didn't stop a huge volume of bad orders from being fired at all the exchanges, even futures exchanges like the CME.
HFT was involved in at least the following ways:
1. Some arbitrage algos went nuts trying to capture prices that weren't there. Most shut themselves down due to risk parameters - volatility, significant differences between venues, quotes crossing the NBBO, etc. Badly written algos no doubt contributed.
2. Because of the seriously odd behavior, most HFT shops pulled out their orders.
3. Due to 2 liquidity dried up, making the volatility worse.
4. It looks like there was feedback caused by algos reacting to delayed price data.
Note: this is all my own analysis - some other observers had published similar findings.
Similar things, although not as bad, have happened a few times in the 6 or so years I've been in HFT.
Fact: studies has concluded that HFT is overall good for the marketplace.
Question: Who commissioned such studies? Are there any other studies commissioned by other people/groups? Do all studies of HFT reach the same conclusion? Are all such studies commissioned by the same people?
Overarching question: What are your sources?
See here, for some studies (not all 100% positive, I admit, but that's to be expected if they are being fair). Most of these studies appear to be independent... but who knows. I can say for sure that many wealthy traders have a low opinion of HFT and would appreciate studies that were critical of HFT.
http://slashdot.org/comments.pl?sid=3033089&cid=40919707
I haven't read any studies that were overall critical of HFT. Some criticize certain behaviors, algos and practices. I have seen some evidence that certain types of practice could be bad for the market. I haven't worked for a company that engages in them, so I have no first hand experience with such things.
So far, all I have is your say-so. Admittedly, that's also most of all I have for SerpentMage (the GP), but he also provides what looks like a real name in his email, and a short bio. I don't even have that for you. In addition, and no offense meant, but a username such as "tolkienfan" does suggest someone with a bent towards fantasy, and when that's all I have to go by, as compared to SerpentMage's claim to be a quant algo writer, then SerpentMage's comments at least appear to be a bit more authoritative.
Cheers,
Fair enough.:)
The fact is, HFT keeps getting very bad press, but without any facts or logical theories to support it. It seems to be that people don't understand it and thus fear it. Hence I sometimes use pitchfork rhetoric. I figure throwing in a few facts and logic wouldn't hurt. Take it however you choose.
When Alice will sell at 100 and Bob will buy at 80. Yeah, I get that.
Alternatively, they can look for a close price that's acceptable.
. . . Well yeah, if you have $100 worth of trinkets but need to sell, you'll take $80, and eat the loss. You get hosed, but you get money when you need it. Or yeah, you can instead offer to sell for $90, and wait to see if anyone takes that sweeter deal. That's about exactly half-way between getting it now and getting what it's worth. That's like saying a healthy alternative is to sell it to me for half it's value.
I get the whole "everything is worth what the buyer/seller will part with" thing, but it ignores urgency. If you value something at $100, if you accept anything less it means you're between a rock and a hard place. You're hand is being forced. You don't want to sell for less than $100, but pressures are making you. That... doesn't seem quite right.
I see your point. I don't see an alternative. If no one is willing to meet your price, should we force the other side? Because of urgency? This goes directly to the time value of money.
But ok ok, I really don't understand why Bob would offer to buy at 110, when there's already someone offering to sell at 100. That's pretty much a perfect example of an irrational market. Which the free market depends upon. Anyone taking advantage of stupid people being irrational is little better than a con artist.
Irrational prices can come from lots of places. E.g. News is announced in Washington. It will affect the price of oil. But oil trades in Chicago. The first to get the news has an advantage. Oil was priced correctly, but price is dependent on time...
They can unwind their position
I don't really get that. You mean... one buys and one sells?
By unwind I mean if they have a long position they sell it all, and if they have a short position they buy it all so they have 0 position at the end.
Now enter HFT. The market maker is out of a job. The spread is now $0.01, and the bid/ask is around 90.10/90.11.
This part. This right here. WHY? How does HFT makes the spread go from 0.25 to 0.01? That is a complete mystery to me.
HFT market makers compete with each other.
Example:
Say we have market maker Carl and Dirk.
Carl quotes 90.05/90.20. He makes almost 15c on each buy+sell pair.
Later Dirk starts trading. He can get trades only if he quotes better prices. So he quotes 90.06/90.19.
Carl quotes 90.07/90.18 and so on.
Note: They can only keep doing this while they are making money. If there were a transaction tax of 10c imposed, Carl could quote no better than 90.04/90.25. That's 10c tax for the buy, 10c tax for the sell and 1c for some profit (and some exchange fees come out). But Dirk can do no better. So the investor bears the brunt of the tax, not the market maker.
PS: Market makers used to make agreements with each other to keep spreads artificially wide. Eventually rules were enacted making the practice illegal, but still...
Even if they didn't have such agreements, they can only do that job while it's profitable. When someone else does a better job of market making you only get the losing trades. E.g. ABC announces a quarterly loss. Some guy, Edward, sees the market maker's bids for 90.25 and 90.26. He sells short getting it at 90.255 on average. Eventually the market maker reacts and moves his quotes down to the new bid/ask of 88.50/88.75. Edward buys back at 88.75. The market maker just took a loss. This happens a lot. Market makers have to be fast to compete and to avoid losses. HFT was inevitable.
Lastly, neither Bob nor Alice can really afford to trade directly on the exchange.
That's exactly the point. Alice and Bob could place limit orders on the exchange and wait. But then THEY ARE PART OF THE SPREAD!
Alternatively, they can look for a close price that's acceptable. Before HFT the best bid would be, say 90 and the best ask would be 90.25. Bob and Alice are about $10 off of their notion of the value, but the market maker makes $0.25. But neither Bob nor Alice have lost $10. They can unwind their position for a total loss of $0.25 per share.
Now enter HFT. The market maker is out of a job. The spread is now $0.01, and the bid/ask is around 90.10/90.11. If Bob buys, then immediately sells, it costs him $0.01.
In all this I've ignored exchange fees - which is reasonable since they are a fraction of $0.01
Lastly, neither Bob nor Alice can really afford to trade directly on the exchange. They really trade through a broker, who buy law cannot charge over the going rate. Say between 0.5% and 1% and a flat fee... That's 45c+ per trade which is 1buy + 1 sell totaling 90c! That's even worse than the market maker in the first place!
I love how plenty of people are ignorant but sure of their opinion. Grab your pitchforks, folks, there's something here we don't understand!
That should get you started. There are plenty of references to the literature for you to follow if you're so inclined.
Some quotes: On net, it is probable HFTs have a positive impact on market quality. (Hirschey 2011) The emergence of HFTs has coincided with a substantial decrease in quoted and realized spreads on exchanges (Castura, Litzenberger, Gorelick, and Dwivedi 2010). In our interpretation, HFTs are traders that make the market extremely ecient, by incorporating information as soon as it becomes available. (Jovanovic and Menkveld 2011)
Nonsense. Crashes have always happened. People see the price drop and want to sell before they lose too much. Lack of liquidity has always contributed, because it means the best bid drops faster, which causes more panic. HFT market making ensures the spread is narrow, and supplies liquidity that stabilizes the market. Insightful my ass.
Slashdot Mirror
You realize you are promoting a routine mutilating surgery for possible (but unproven) reduction in the likelihood of various unlikely diseases?
Other countries in the world look at the US and laugh.
That article doesn't really support the AAPs position...
Circumcision causes disfigurement and death every year.
Admittedly, I haven't rtfa (this IS slashdot), but I'd suggest that there is likely a causative factor in the kinds of parents that choose to circumcise in those parts of the world.
I can say for sure that I get much of my sexual stimulation from the foreskin itself. Denying that of someone because of possible issues later is ridiculous. It's also against the Hippocratic "do no harm" tenet.
Pediatrics is the only branch of medicine that routinely performs unnecessary surgeries.
The US is almost alone in routinely performing circumcision for non-religious reasons.
Canadian pediatrics recommends against it, IIRC.
And every year circumcisions cause disfigurement and death.
But hey, its a money maker!
You count to do multiplication?? Really?
What about powers?
How do you even know what to do to get to the result? The procedures must be memorized. The symbols. Shortcuts.
If you've ever worked with kids you'll know that you can't merely teach a method and let them go. They get stuck at the first step. They have to learn facts along with the method.
How fast can you count the first billion natural numbers? Time yourself... I'm curious.
You do need to memorize some results. ...and then repeat that process to do multiplication? Most people will look at 5*5 and recall 25 without going thru a tedious counting process. Ymmv
You don't, for example, do repeated increments in your head to add, do you?
Educators probably rely a little to much on rote learning, but a certain amount is crucial.
They're using a religious exemption, Skippy.
IMO that exemption shouldn't exist.
Smallpox killed 300-500 million people, during the 20th century.
It's not a scam.
That sounds like a rhetorical question, but if you think an unvaccinated person cannot make vaccinated people sick, then you are seriously ignorant.
Please go learn about vacination, the various vaccines in common usage, their efficacy, and the diseases they help prevent.
This is way too serious an issue to go making this kind of comment without actually knowing anything about the subjevt matter.
Oh there are plenty of Jews and Christians that want the same thing for the US.
I'm going to steal that.
Brilliant!
If we don't vote for a lizard, the wring lizard might get in.
When people refer to "calculus" without a qualifier it almost always means "differential and integral calculus". Lambda calculus is a different beast.
I've had the pleasure of fitting models to some interesting data. This required non-linear fitting... differentiation and linear algebra.
Fun stuff!
HFT firms have no special privilege - they pay the same exchange fees that all members pay.
It's true that some market makers pull or widen significantly before certain announcements. But energy swings so violently after news that they can't afford to stay in there. It's only for a few minutes.
Without HFT the market makers and arbitrageurs would be there anyway. And yes, actually, even flesh and blood market makers would widen the spreads leading up to news - otherwise they would lose money. And I don't mean a little during the swing, I mean overall the losses would outway any profit. Market makers are just ad focused on avoiding losing money from big moves as they are focused on making a tiny profit on the spread.
You are sure that HFT MUST be bad, but you don't understand it, and frankly, the evidence is against you.
All orders are attributed to a specific trader. Although an algo selected when to order, what price and so on, the trader takes responsibility for every order.
And it actually is enforced. I have first hand knowledge of inquiries and punitive actions.
Behaviors that are considered manipulative are punished very severely. They aren't good for the exchange, SRO or marketplace.
Fact: the biggest crashes over the past couple of years have occurred when HFT pulled out due to bad information and impossible quotes.
Question: Where did those impossible quotes come from? Why were they impossible? What prompted such quotes? Were the algo quotes, or human quotes? Was HFT any significant factor in why such impossible quotes were made?
The quotes I'm referring to were market data from the 2010 flash crash incident. That market data was published by NYSE. The prices were correct, but the data had been delayed. But the bad part was the timestamps on the data indicated they were fresh. It appears that NYSE queued up market data, but then timestamped it on exit from the queue. This created the appearance of huge arbitration opportunities between NYSE and the other equities venues. Those opportunities really didn't exist, but that didn't stop a huge volume of bad orders from being fired at all the exchanges, even futures exchanges like the CME.
HFT was involved in at least the following ways:
1. Some arbitrage algos went nuts trying to capture prices that weren't there. Most shut themselves down due to risk parameters - volatility, significant differences between venues, quotes crossing the NBBO, etc. Badly written algos no doubt contributed.
2. Because of the seriously odd behavior, most HFT shops pulled out their orders.
3. Due to 2 liquidity dried up, making the volatility worse.
4. It looks like there was feedback caused by algos reacting to delayed price data.
Note: this is all my own analysis - some other observers had published similar findings.
Similar things, although not as bad, have happened a few times in the 6 or so years I've been in HFT.
Fact: studies has concluded that HFT is overall good for the marketplace.
Question: Who commissioned such studies? Are there any other studies commissioned by other people/groups? Do all studies of HFT reach the same conclusion? Are all such studies commissioned by the same people?
Overarching question: What are your sources?
See here, for some studies (not all 100% positive, I admit, but that's to be expected if they are being fair). Most of these studies appear to be independent... but who knows. I can say for sure that many wealthy traders have a low opinion of HFT and would appreciate studies that were critical of HFT. http://slashdot.org/comments.pl?sid=3033089&cid=40919707
I haven't read any studies that were overall critical of HFT. Some criticize certain behaviors, algos and practices. I have seen some evidence that certain types of practice could be bad for the market. I haven't worked for a company that engages in them, so I have no first hand experience with such things.
So far, all I have is your say-so. Admittedly, that's also most of all I have for SerpentMage (the GP), but he also provides what looks like a real name in his email, and a short bio. I don't even have that for you. In addition, and no offense meant, but a username such as "tolkienfan" does suggest someone with a bent towards fantasy, and when that's all I have to go by, as compared to SerpentMage's claim to be a quant algo writer, then SerpentMage's comments at least appear to be a bit more authoritative.
Cheers,
Fair enough. :)
The fact is, HFT keeps getting very bad press, but without any facts or logical theories to support it. It seems to be that people don't understand it and thus fear it. Hence I sometimes use pitchfork rhetoric. I figure throwing in a few facts and logic wouldn't hurt. Take it however you choose.
But then THEY ARE PART OF THE SPREAD!
When Alice will sell at 100 and Bob will buy at 80. Yeah, I get that.
Alternatively, they can look for a close price that's acceptable.
. . . Well yeah, if you have $100 worth of trinkets but need to sell, you'll take $80, and eat the loss. You get hosed, but you get money when you need it. Or yeah, you can instead offer to sell for $90, and wait to see if anyone takes that sweeter deal. That's about exactly half-way between getting it now and getting what it's worth. That's like saying a healthy alternative is to sell it to me for half it's value. I get the whole "everything is worth what the buyer/seller will part with" thing, but it ignores urgency. If you value something at $100, if you accept anything less it means you're between a rock and a hard place. You're hand is being forced. You don't want to sell for less than $100, but pressures are making you. That... doesn't seem quite right.
I see your point. I don't see an alternative. If no one is willing to meet your price, should we force the other side? Because of urgency? This goes directly to the time value of money.
But ok ok, I really don't understand why Bob would offer to buy at 110, when there's already someone offering to sell at 100. That's pretty much a perfect example of an irrational market. Which the free market depends upon. Anyone taking advantage of stupid people being irrational is little better than a con artist.
Irrational prices can come from lots of places. E.g. News is announced in Washington. It will affect the price of oil. But oil trades in Chicago. The first to get the news has an advantage. Oil was priced correctly, but price is dependent on time...
They can unwind their position
I don't really get that. You mean... one buys and one sells?
By unwind I mean if they have a long position they sell it all, and if they have a short position they buy it all so they have 0 position at the end.
Now enter HFT. The market maker is out of a job. The spread is now $0.01, and the bid/ask is around 90.10/90.11.
This part. This right here. WHY? How does HFT makes the spread go from 0.25 to 0.01? That is a complete mystery to me.
HFT market makers compete with each other.
Example:
Say we have market maker Carl and Dirk.
Carl quotes 90.05/90.20. He makes almost 15c on each buy+sell pair.
Later Dirk starts trading. He can get trades only if he quotes better prices. So he quotes 90.06/90.19.
Carl quotes 90.07/90.18 and so on.
Note: They can only keep doing this while they are making money. If there were a transaction tax of 10c imposed, Carl could quote no better than 90.04/90.25. That's 10c tax for the buy, 10c tax for the sell and 1c for some profit (and some exchange fees come out). But Dirk can do no better. So the investor bears the brunt of the tax, not the market maker.
PS: Market makers used to make agreements with each other to keep spreads artificially wide. Eventually rules were enacted making the practice illegal, but still... Even if they didn't have such agreements, they can only do that job while it's profitable. When someone else does a better job of market making you only get the losing trades. E.g. ABC announces a quarterly loss. Some guy, Edward, sees the market maker's bids for 90.25 and 90.26. He sells short getting it at 90.255 on average. Eventually the market maker reacts and moves his quotes down to the new bid/ask of 88.50/88.75. Edward buys back at 88.75. The market maker just took a loss. This happens a lot. Market makers have to be fast to compete and to avoid losses. HFT was inevitable.
Lastly, neither Bob nor Alice can really afford to trade directly on the exchange.
Yeah, I hear you about the
That's exactly the point.
Alice and Bob could place limit orders on the exchange and wait.
But then THEY ARE PART OF THE SPREAD!
Alternatively, they can look for a close price that's acceptable.
Before HFT the best bid would be, say 90 and the best ask would be 90.25. Bob and Alice are about $10 off of their notion of the value, but the market maker makes $0.25. But neither Bob nor Alice have lost $10. They can unwind their position for a total loss of $0.25 per share.
Now enter HFT. The market maker is out of a job. The spread is now $0.01, and the bid/ask is around 90.10/90.11.
If Bob buys, then immediately sells, it costs him $0.01.
In all this I've ignored exchange fees - which is reasonable since they are a fraction of $0.01
Lastly, neither Bob nor Alice can really afford to trade directly on the exchange. They really trade through a broker, who buy law cannot charge over the going rate. Say between 0.5% and 1% and a flat fee... That's 45c+ per trade which is 1buy + 1 sell totaling 90c! That's even worse than the market maker in the first place!
I love how plenty of people are ignorant but sure of their opinion. Grab your pitchforks, folks, there's something here we don't understand!
There is some criticism in here too...
http://papers.ssrn.com/sol3/papers.cfm?abstract_id=1712765
http://www.efinancialnews.com/story/2011-08-16/capital-markets-cooperative-research-hft
http://faculty.haas.berkeley.edu/hender/HFT-PD.pdf
https://www2.bc.edu/~taillard/Seminar_spring_2012_files/Hirschey.pdf
That should get you started. There are plenty of references to the literature for you to follow if you're so inclined.
Some quotes:
On net, it is probable HFTs have a positive impact on market quality. (Hirschey 2011)
The emergence of HFTs has coincided with a substantial decrease in quoted and realized spreads on exchanges (Castura, Litzenberger, Gorelick, and Dwivedi 2010).
In our interpretation, HFTs are traders that make the market extremely ecient, by incorporating information as soon as it becomes available. (Jovanovic and Menkveld 2011)
Nonsense.
Crashes have always happened. People see the price drop and want to sell before they lose too much. Lack of liquidity has always contributed, because it means the best bid drops faster, which causes more panic. HFT market making ensures the spread is narrow, and supplies liquidity that stabilizes the market.
Insightful my ass.
Studies show that HFT helps the marketplace as a whole.