Its probably too late to comment on this, but just a minor point; A company's valuation determines its ability to access the credit markets. A higher stock price means the company is probably going to source debt at better rates than a similar company with a lower stock price. Company's do care about stock price.
Not sure if HFT will buy all the product on all the store shelves on the off chance that you are going to reach for another unit and might be willing to pay up for the next one. Too much risk, very little reward.
What if the store just restocked everything instantly? (big seller). Whats the HFT guy going to do with all that product he just bought?
Now imagine a scenario where its not a supermarket but a open-outcry farmers market. Farmer A just sold tomatoes for X. Farmer B wants to be the next guy to sell tomatoes for X. He needs to have the fastest response time to be able to yell "tomatoes for X" before anyone else does. Hence the low latency connections.
I don't believe above represents an arbitrage scenario. That would only work if the trader in the middle "saw" these orders before they hit the order book. The way it would work in an electronic market is as follows:
A enters order to sell for X. IF the order is not immediately bought, it enters the orderbook and potentially alters the national best offer to X.
B enters order to buy at Y=X+C. If offer X still exists, no matter where Y is entered, it will have to be routed to the exchange where X is and trade through X before it can trade at X+0.01, 0.02, 0.03 <= C. This is a rule.
High frequency liquidity providing traders that want to sell at X along with the first order will use their colocation and other speed advantages to attempt be the first order to sell at X after the original order. Other (liquidity taking) algorithms might anticipate a Y buy and buy up X beforehand to sell at Y. That is not arbitrage, its buying low and selling high.
India actually happens to be among the handful of countries in the world where consumption of marijuana is legal. It's the *only* country in the world where such consumption is sanctioned by the government. You can walk up to a govt. run store, hand the dude some cash and walk away with a bag of reefer.... good times...
This is almost certainly not the source code in question. You are talking about a standard vwap engine that some firms use to trickle trade large orders. In fact only an idiot pension fund manager would ever call up his broker and tell him to sell 1MM shs of anything. ( or maybe one that's expecting ringside seats to the superbowl in the mail ). Anyway, what used to happen in the "ye olde days" to cop a phrase from another poster is that said idiot fund manager would call up 4-5 brokers at various banks and ask them to price a block of 1MM shares. If you did not consummate the trade with the broker thats on the phone right away, he starts dumping ( or buying ) the instant you get off the phone ( sometimes while he keeps you on the phone ). So that no matter what strategy you used to break up the order, the damage is done much before your flow even comes on the market.
The software in question appears to be some sort of arb engine. Stat arb, index arb or futures arb. These engines typically need pricing data from various markets and execute trades on all those markets. They run a very small shortest path loop with one anchor to figure out whether certain items are getting out of sync ( they get out of sync for example, when a regular joe like you or me leaves a limit order hanging out there waiting for the market to trade through our price ). An engine like this will try to do a closed loop trade that makes maybe a penny or two with 0 risk at the end; ie there is not a great expectation of profit per trade but thousands of these an hour and millions a day add up pretty quickly. Any strategies layered on top of these ( wavelet anaylysis, momentum etc ) will try to increase the 1-2c to 2-5c etc. Based on the stuff this engine was touching and reading the resume on what the dude was doing at GS, this is my best guess.
Delhi went all CNG for public vehicles a long time ago. Smog is no longer a big issue in inner delhi.
Mumbai and Pune are in the process of converting all public vehicles to CNG. Many private taxi operators have already converted.
While elecric energy is a nice solution for city vehicles, I think this is merely shifting the pollution from cities to places where coal or furnace oil is burnt to produce electricity. I know that this can be state controlled etc, but with more and more electricity being supplied to vehicular needs, what's to stop us from getting into a situation where a power plant is "too big to fail"...
I believe that CNG is a nice compromise in that, there is still plenty of it around, it's just as easy to store, transport, distribute and consume as oil and burns more cleanly. New Delhi is a fine example of the positive impact of predominantly CNG burning vehicles. I wonder why we haven't seen some real effort behind pushing CNG vehicles here in the US. In fact, current petrol vehicles can be converted to burn CNG with a minor one-time modification.
... As to why the visa system is clogged... Maybe the economic hard times have hit government offices partially responsible for it as well? Oh, what sweet revenge. -_-
I know you are kidding, but the visa system has always been clogged. So I would say the government continues to operate at the same level of efficiency as before.:-)
No, at least while I was a grad student, an (F1) student visa was only able to get you a 1 year apprenticeship after graduation. No way to apply for residency.
If the company you apprentice with for a year decides to keep you on, they sponsor an H1B visa that allows you to apply for residency.
The process of actually getting residency is horribly convoluted and takes between 5 and 8 years to complete. Not to mention all the shyster lawyers out there (yechh).
Funding opportunities for foreign students come from university resources only. They are not eligible for FAFSA, state or local government scholarships.
Fellowships are rare, most foreign graduate students end up getting a teaching assistantship or research assistantship. Undergrads don't even have that. Further, they cannot take up jobs off-campus which generally pay as much or better than on-campus jobs.
From the top down, if the banks don't have cash (or easy access to source cash intrabank, what is really drying up, the LIBOR is as high as it has ever been), they cannot lend it.
So lets fix that problem.
Banks need money for credit lines? Lets figure out a way to get them that cash.
ATMs in danger of running out of cash? Lets figure out a way to keep them full.
There is no reason for such a broad brush $700 billion "maturity value" purchase of this toxic paper. Oh BTW... do you know that buying the bonds at "maturity value" is the same as giving these banks an indefinite interest free loan?
There is a lot of talk about how the taxpayer ( I hate to be referred to like that ) is going to make a little money... BULLSHIT!!!... the taxpayer is going to have to wait a long long time before the underlying assets recover and these bonds can be written up to reflect an increase in the value of the underlying collateral ( the homes ). And fat chance that these banks are going to pay maturity value on these bonds to buy them back once the credit crunch has passed. A better solution would be to outright put money into the banks, through preferred stocks or warrants, yes it would be massively dilutional and probably wipe out the current shareholders, but at least this way the taxpayer gets the benefit of future dividends and stock price appreciation. In the current plan, we get nothing.
Laser beams would not be dashed lines like that would they? Most likely they would look like an infinitely long straight beam of light shooting off into space.
The image might be showing plasma based weapons that are essentially firing mass at high velocities and parabolic paths are acceptable.
offtopic??? C'mon mods. really!!!. Can you find a better example of piracy in the current day? What's wrong with pointing out that privateerism is alive and well even in the 21st century.
I mean... avast, haul the quartermast, trim the mainsail, drop your cocks and grab your socks... time to go a-plunderin'.
I have no problem with prosecuing the naked short selling rules. What's at issue is that shorting shares ( naked or otherwise ) has been banned on all financials.
What's incredible is that companies like Sears, Netflix, Capstone have been crying foul for years and wall st laughed and kept shorting them blind while the SEC looked the other way. Here is an example. The rule was never enforced. Today wall st is in the crosshairs and everyone is running around trying to prevent short sales.
These markets are not orderly. An entrenched well-connected few are pulling every string in sight to protect their interests and we all are going to end up footing the bill when the piper comes calling.
Today is also trade like a pirate day. Since the administration has just ruled that all the shorts should just hand over money to the longs.
This is the most egregious example of willful market manipulation since 1929. We have ringside seats to the greatest act of sanctioned piracy since the french "blessed" william kidd.
Many brokers support FIX, so you can use quickfix to interact with major brokerage firms.
Many data vendors supply APIs that will let you create an adapter in python/C/C++ or Java to pull prices into gnucash if thats your preferred analysis tool.
There are table widgets that you can buy to cobble together a FOSS charting/graphing package using opengl. Just google table widget, chart package etc.
AFAIK, there is no end-to-end foss solution that will perform like a trader GUI.
No, arbitrage is a necessary component of orderly markets. A perfectly balanced market has 0 arbitrage opportunity.
Example: ETF Conversion/Redemption. The ETF is priced in real-time based on the price of its components. Fast systems are able to detect small discrepancies in the price of some of the components and the basket as a whole and are able to execute trades ( say BUY ) on the components and the contra trade ( SELL ) on the ETF and then redeem the ETF from the components to pair off the contra trade.
This arbitrage always works to keep the components perfectly in line with the ETF itself. I think that is a good thing.
I too am blown away by that number. My first instinct was that 3 ms is the round trip time from their edge servers to their matching engine; to receive orders, produce a match and send out confirm messages.
Some SAN systems though, will write data to RAM and return a write commit back to the host and write the data to disk offline. That might be possible in under 3 ms. I don't know for sure.
Its scary how close to the truth parent is. I was working with our Telco provider once, debugging some latency I was seeing on our dedicated Internet circuit. A BC came up to me and suggested we use cable modem because its cheaper and he gets fantastic download speeds at home.
Unfortunately, I have to agree. This is mostly true simply because of the expedience of having to fix something, or add something and roll it out so that the trader or the desk can start making money right away.
Most places that I have worked, the average lifespan of an IT manager/developer etc working on something any more advanced than cookie cutter java web apps is under 3 years. Projects fail, traders skip jobs, desks are reshuffled, priorities are changed on a near daily basis... Not the ideal environment to setup a "real" IT shop.
LSE boo-boo looks more like bad/inadequate planning and design than anything else. I dont know if.Net is the culprit here, I dont see how it could be. It is simply a technology platform and from what I know about it, it does what it says it does and it is possible to get a good estimate of how the platform will behave under "real" conditions. I find it hard to blame the technology for this fiasco, this was human failure, not failure of tech.
His strength left him because of a prophesy I think... Something about the next guardian. I am reasonably certain this was not a case of failed dorsiflexors.
I am sort-of in agreement with you wrt upper management. You have to remember though that a very significant portion of the current crop of upper management is the old guard that operated with nothing more than a phone and a typewriter as IT resources.
To them, IT will never be more than a necessary evil, something that needs to exist ( like the janitorial staff ), but is not core to the success of the enterprise. Look at most of the businesses around you, how has IT changed anything about how things get done except make stuff that was already mundane and tedious faster and more automated?
Its not until enterprises start to take a fresh look at the way they do business and incorporate technology as a core element that we will start to see significant investment in internally developed IT assets. Some stuff is still going to be outsourced, there is no way around that, but the critical stuff will tend to stay local and should provide plenty of opportunity to the new crop of developers
Its probably too late to comment on this, but just a minor point; A company's valuation determines its ability to access the credit markets. A higher stock price means the company is probably going to source debt at better rates than a similar company with a lower stock price. Company's do care about stock price.
Not sure if HFT will buy all the product on all the store shelves on the off chance that you are going to reach for another unit and might be willing to pay up for the next one. Too much risk, very little reward.
What if the store just restocked everything instantly? (big seller). Whats the HFT guy going to do with all that product he just bought?
Now imagine a scenario where its not a supermarket but a open-outcry farmers market. Farmer A just sold tomatoes for X. Farmer B wants to be the next guy to sell tomatoes for X. He needs to have the fastest response time to be able to yell "tomatoes for X" before anyone else does. Hence the low latency connections.
I don't believe above represents an arbitrage scenario. That would only work if the trader in the middle "saw" these orders before they hit the order book. The way it would work in an electronic market is as follows:
High frequency liquidity providing traders that want to sell at X along with the first order will use their colocation and other speed advantages to attempt be the first order to sell at X after the original order. Other (liquidity taking) algorithms might anticipate a Y buy and buy up X beforehand to sell at Y. That is not arbitrage, its buying low and selling high.
hehe.
India actually happens to be among the handful of countries in the world where consumption of marijuana is legal. It's the *only* country in the world where such consumption is sanctioned by the government. You can walk up to a govt. run store, hand the dude some cash and walk away with a bag of reefer.... good times...
This is almost certainly not the source code in question. You are talking about a standard vwap engine that some firms use to trickle trade large orders. In fact only an idiot pension fund manager would ever call up his broker and tell him to sell 1MM shs of anything. ( or maybe one that's expecting ringside seats to the superbowl in the mail ). Anyway, what used to happen in the "ye olde days" to cop a phrase from another poster is that said idiot fund manager would call up 4-5 brokers at various banks and ask them to price a block of 1MM shares. If you did not consummate the trade with the broker thats on the phone right away, he starts dumping ( or buying ) the instant you get off the phone ( sometimes while he keeps you on the phone ). So that no matter what strategy you used to break up the order, the damage is done much before your flow even comes on the market.
The software in question appears to be some sort of arb engine. Stat arb, index arb or futures arb. These engines typically need pricing data from various markets and execute trades on all those markets. They run a very small shortest path loop with one anchor to figure out whether certain items are getting out of sync ( they get out of sync for example, when a regular joe like you or me leaves a limit order hanging out there waiting for the market to trade through our price ). An engine like this will try to do a closed loop trade that makes maybe a penny or two with 0 risk at the end; ie there is not a great expectation of profit per trade but thousands of these an hour and millions a day add up pretty quickly. Any strategies layered on top of these ( wavelet anaylysis, momentum etc ) will try to increase the 1-2c to 2-5c etc. Based on the stuff this engine was touching and reading the resume on what the dude was doing at GS, this is my best guess.
I think that was the destination code, spelt a little badly.
Delhi went all CNG for public vehicles a long time ago. Smog is no longer a big issue in inner delhi.
Mumbai and Pune are in the process of converting all public vehicles to CNG. Many private taxi operators have already converted.
While elecric energy is a nice solution for city vehicles, I think this is merely shifting the pollution from cities to places where coal or furnace oil is burnt to produce electricity. I know that this can be state controlled etc, but with more and more electricity being supplied to vehicular needs, what's to stop us from getting into a situation where a power plant is "too big to fail"...
I believe that CNG is a nice compromise in that, there is still plenty of it around, it's just as easy to store, transport, distribute and consume as oil and burns more cleanly. New Delhi is a fine example of the positive impact of predominantly CNG burning vehicles. I wonder why we haven't seen some real effort behind pushing CNG vehicles here in the US. In fact, current petrol vehicles can be converted to burn CNG with a minor one-time modification.
ayy... I beg to differ. Tony sleeps with the fishes, I made sure of that. Now, where's that canoli?
... As to why the visa system is clogged... Maybe the economic hard times have hit government offices partially responsible for it as well? Oh, what sweet revenge. -_-
I know you are kidding, but the visa system has always been clogged. So I would say the government continues to operate at the same level of efficiency as before. :-)
No, at least while I was a grad student, an (F1) student visa was only able to get you a 1 year apprenticeship after graduation. No way to apply for residency.
If the company you apprentice with for a year decides to keep you on, they sponsor an H1B visa that allows you to apply for residency.
The process of actually getting residency is horribly convoluted and takes between 5 and 8 years to complete. Not to mention all the shyster lawyers out there (yechh).
Funding opportunities for foreign students come from university resources only. They are not eligible for FAFSA, state or local government scholarships.
Fellowships are rare, most foreign graduate students end up getting a teaching assistantship or research assistantship. Undergrads don't even have that. Further, they cannot take up jobs off-campus which generally pay as much or better than on-campus jobs.
From the top down, if the banks don't have cash (or easy access to source cash intrabank, what is really drying up, the LIBOR is as high as it has ever been), they cannot lend it.
So lets fix that problem.
There is no reason for such a broad brush $700 billion "maturity value" purchase of this toxic paper. Oh BTW... do you know that buying the bonds at "maturity value" is the same as giving these banks an indefinite interest free loan?
There is a lot of talk about how the taxpayer ( I hate to be referred to like that ) is going to make a little money... BULLSHIT!!!... the taxpayer is going to have to wait a long long time before the underlying assets recover and these bonds can be written up to reflect an increase in the value of the underlying collateral ( the homes ). And fat chance that these banks are going to pay maturity value on these bonds to buy them back once the credit crunch has passed. A better solution would be to outright put money into the banks, through preferred stocks or warrants, yes it would be massively dilutional and probably wipe out the current shareholders, but at least this way the taxpayer gets the benefit of future dividends and stock price appreciation. In the current plan, we get nothing.
Laser beams would not be dashed lines like that would they? Most likely they would look like an infinitely long straight beam of light shooting off into space.
The image might be showing plasma based weapons that are essentially firing mass at high velocities and parabolic paths are acceptable.
There are 2 NASA payloads selected for the Chandrayaan I, not one.
offtopic??? C'mon mods. really!!!. Can you find a better example of piracy in the current day? What's wrong with pointing out that privateerism is alive and well even in the 21st century.
I mean... avast, haul the quartermast, trim the mainsail, drop your cocks and grab your socks... time to go a-plunderin'.
I have no problem with prosecuing the naked short selling rules. What's at issue is that shorting shares ( naked or otherwise ) has been banned on all financials.
What's incredible is that companies like Sears, Netflix, Capstone have been crying foul for years and wall st laughed and kept shorting them blind while the SEC looked the other way. Here is an example. The rule was never enforced. Today wall st is in the crosshairs and everyone is running around trying to prevent short sales.
These markets are not orderly. An entrenched well-connected few are pulling every string in sight to protect their interests and we all are going to end up footing the bill when the piper comes calling.
Today is also trade like a pirate day. Since the administration has just ruled that all the shorts should just hand over money to the longs.
This is the most egregious example of willful market manipulation since 1929. We have ringside seats to the greatest act of sanctioned piracy since the french "blessed" william kidd.
Oh and... Arrrh.
AFAIK, there is no end-to-end foss solution that will perform like a trader GUI.
No, arbitrage is a necessary component of orderly markets. A perfectly balanced market has 0 arbitrage opportunity.
Example: ETF Conversion/Redemption. The ETF is priced in real-time based on the price of its components. Fast systems are able to detect small discrepancies in the price of some of the components and the basket as a whole and are able to execute trades ( say BUY ) on the components and the contra trade ( SELL ) on the ETF and then redeem the ETF from the components to pair off the contra trade.
This arbitrage always works to keep the components perfectly in line with the ETF itself. I think that is a good thing.
I too am blown away by that number. My first instinct was that 3 ms is the round trip time from their edge servers to their matching engine; to receive orders, produce a match and send out confirm messages.
Some SAN systems though, will write data to RAM and return a write commit back to the host and write the data to disk offline. That might be possible in under 3 ms. I don't know for sure.
Its scary how close to the truth parent is. I was working with our Telco provider once, debugging some latency I was seeing on our dedicated Internet circuit. A BC came up to me and suggested we use cable modem because its cheaper and he gets fantastic download speeds at home.
Unfortunately, I have to agree. This is mostly true simply because of the expedience of having to fix something, or add something and roll it out so that the trader or the desk can start making money right away.
Most places that I have worked, the average lifespan of an IT manager/developer etc working on something any more advanced than cookie cutter java web apps is under 3 years. Projects fail, traders skip jobs, desks are reshuffled, priorities are changed on a near daily basis... Not the ideal environment to setup a "real" IT shop.
LSE boo-boo looks more like bad/inadequate planning and design than anything else. I dont know if .Net is the culprit here, I dont see how it could be. It is simply a technology platform and from what I know about it, it does what it says it does and it is possible to get a good estimate of how the platform will behave under "real" conditions. I find it hard to blame the technology for this fiasco, this was human failure, not failure of tech.
I could not agree with you more, especially on the going back to C bit.
After 6+ years in C++, I have realized that C is *just right* for certain libraries.
I have not stopped coding C++, but I am slowly moving to C on the lower level performance critical portions of my libararies.
His strength left him because of a prophesy I think... Something about the next guardian. I am reasonably certain this was not a case of failed dorsiflexors.
I am sort-of in agreement with you wrt upper management. You have to remember though that a very significant portion of the current crop of upper management is the old guard that operated with nothing more than a phone and a typewriter as IT resources.
To them, IT will never be more than a necessary evil, something that needs to exist ( like the janitorial staff ), but is not core to the success of the enterprise. Look at most of the businesses around you, how has IT changed anything about how things get done except make stuff that was already mundane and tedious faster and more automated?
Its not until enterprises start to take a fresh look at the way they do business and incorporate technology as a core element that we will start to see significant investment in internally developed IT assets. Some stuff is still going to be outsourced, there is no way around that, but the critical stuff will tend to stay local and should provide plenty of opportunity to the new crop of developers