i'm ali velshi "on target", man versus machine, see how the need for speed turned america's volatile stock market upside down. revenge of the nerds, meet the brains behind the computers that can make hundreds of million in milliseconds this last month has been a challenging and turbulent time for investigators in the stock market. we see them going up or down. if you trade stocks, or ira, it's times like these, that you can try your sole. i want to talk to you about what might, might be feeding all of the volatility. to do that, i want to draw a comparison to global warming.

this assumes we agree that global warming is happening. some don't. if you believe the planet is heating up because of human activity. we'll agree with scientists that say global warming is causing greater weather volatility, blizzards, tif onlies. critics say the stock market is brewing up its own storm because of traders, selling larger volumes of shares faster and faster. critics point to high frequency trading. countless trades conducted in split-second intervals. tens of billions conducted every day. measured in fractions of a pay. high-volume speed trading will heat up markets, making for a mark et filled with hurricanes, blizzards and typhoons. if you stay in it you have to be

prepared to ride out the storm. it speaks volumes about the transformation that stock exchanges have undergone in years. as i reported last year, it's because of the need for speed has become the holy grail for market traders. >> back on september 5th to 13, something remarkable happened at the closing well. on that day. a company called ultra salon cosmetics raised a report, it fell short of expectations, mata data firm said high frequency trading computers sold 14,000 shares of stock, at $121 each, totalling 1.6 million worth of stock. at 4:00p.m. the stock closed at $118 a share. losing $3 in less than a second. in that rapid sell off.

the stock, within the one second after 4:00p.m. they estimated that high frequency traders could have pack eth $28,000. a the -- pok eted $28,000. a feat only they could have pulled off, and it was illegal. high frequency traders operating near the speed of light turned what happened on the floor of the new york stock exchange into little more. look behind the curtain, and you see why humans can't compete. >> the only participants in the time frame are these type of machines. humans aren't act perceive information, to act on it, in that time frame. the founder is a high frequency trading. he's been watching it evolve. growing to hundreds of firms. from new york and chicago. it's not measured in seconds,

milliseconds. it's one second chopped up into a thousand parts. it takes 200 milliseconds to recognise a written word. do you know how long it takes a high speed trader to trade. forget about investing fundamentals. or diversification. these days, if you are not trading at those speeds, you are toast. it's not just speed. high frequency traders have the special sauce. secret sauce is a black book algorithm, costing half a million to $10 millions each. they scour the market for queue -- cues, including stock prices moving higher or lower, changes in trade or a disappointing earnings report. all are in the blink of an eye.

tens of thousands times a day, generating profits of a fraction of a penny per trade. it adds up. >> if you have an algorithmic code that does that. literally hundreds of thousands of time per day, it takes money. high frequency firms, some that bragged in the test. >> high frequency trading setting prices for everything that traded. stocks, bonds, futures, commodities. it affects everything. orange juice to the price of mortgages, and some think there's nothing to worry about. like jim, the former chief from the securities and exchange commission. are the markets broken. i would say not. i would say that they had many of the same issues arise, the same thing when computers or

telephone from the trading area. for mum and pop, focussing on the long term, being an investors, you are okay others are calling for more regulation, and say high frequency traders have an unfair advantage. regulators are investigating the relationship with exchanges. bart chilton, commissioner of the traiting commission says traders that use high frequency trade programs are cheaters, saying they are running wild in the parked. >> technology is great, we love it. i want it for the markets. currently there's nos rules or regulations on chesters. i'm concerned that the high frequency traders may royal the market. regulators say high frequency

trading. the dow jones average fell 1,000 points in five minutes. briefly erasing a trillion dollars in market value. only to recover a few minutes later. on april 23rd, 2013, high frequency trading reacted to a hack at the twitter account: it said president barack obama was injured. it erased millions worth of value before it rebound the. the real fear is the market will not recover from a flash cash. and hundreds of millions will be wiped out joining me is the author of high frequency trading. they say aggressive high frequency trading is balanced and good for the markets. the sell off in the market

preceded high frequency trading. you heard my - almost an indictment of some of the parts of high frequency trading. i take it you take a different position than i do. in broad strokes, why? >> i liked the analogy about climate change, and now there's cove si whether it exists or not. if the records are not kept accurate enough over the past decade to ascertain whether there's climate change now tore not. it's not dissimilar to what is happening in financial markets. >> if you look at 20 years ago, the only quotes available to investors were daily open, high, low, and closing prices. the following day, now we are inundated with that. every investor can track online a full-time close. our firm did analysis on flash

crashes and whether using whatever data available, and we found that the flash crashes to date are not more frequent than they were 20, 30, 40 years ago. >> you used to work at the stock exchange as a reporter back in the days it had three floors, it was an open outcry system, before trading by computers. when stock went down some humans had to sell or buy that stock from another human. you certainly didn't have market wide flash crashes in those days, because they - you did. >> there was intervention, more dell tails than the viewers need, but what were called specialists, whose responsibility it was to buy up stocks. that were not days where the market went down by three, four, five, six, 7% and then bounced back and no one knew what

happened. >> they were. they were not reported in the press. >> i was on the floor. i never witnessed such a thing. on the floor you saw the booths where the training happened. it never happened. you never saw things go from green to red and not come back. >> that's the thing about flash crafts. what happened with flashes. they go down low, and they come back. if you look at the crash of may 6, 2010. investors stuck it out, and didn't sell during the panic. they were dubbed better at the end of the day. the prices went up. we measure flash crashes, taking the newspapers, going back 40 years for every stock out there, and compare it to the entry day low to the open. you'll be surprised how many flash crashes there are. now, the reason why you never see it exceed 5%, all the

exchanges have a built in system where if the price drops by 5%. it halts trading that security. >> that's recently been reduced. it was 10%, it's now 7 and 20. you can get to 7% before anyone does anything. >> yes. but the bottom line. they were extremely common going back 10, 20 years ago. they are the only one period in the history of the financial markets. and in front of the 1990ing when there was an unusually low number of flash crashes. few flash crashes for that period of time. if you look back, to the '80s, '70s. to 2000. the number is high. how would you respond to critics, that say that the big swings that you looked at. the big market swings that are

generally speaking sentiment driven. are amplified by high frequency, there's a situation where the sentiment happened overnight in china. in the case of the dow, high frequency trading. they looked for sentiment. follows it down. giving it greater velocity. >> in most crashes, high frequency follow the institutional lead. pension funds, hedge funds and other large players in the markets. those that take short-term positions, they have longer term views, they are the ones that nirfeiate the pressure. and high frequency trades respond to the kind of pressures in the market. continuing to trade around it. >> you think this is what happened in these moves that we saw in august. it was triggered by instituti

institutional sellers. >> that's the conclusion. again, the power of computerized traders is in data analysis. so you are right. people can't blink, people can capture. 24 frames per second. >> this goes a lot faster. >> what a great conversation. the managing director. she's the author, a practical guide to algorithmic strategies. the keys to success in the market. speed and locationation for a first hand look. i didn't go to wall street. i had to go to jersey. i'll take you through next.

there's an arms race in the fibre optic world, and fibre optic cable is a choice. you won't believe the amount of money high frequency companies are willing to spend on cables. they know the key to profit is speed. as i show you last year, these my frequency traders are ready to pay for the privilege of being first in line. back in the 1920s, you'd have to wait about 20 minutes at a ticker tape to see the latest information about a stock. now creating information is almost instantaneous.

travelling at the speed of light. it's so fast only the fastest computers can keep up with the pace. this is the trading floor of the new york stock exchange. it's almost completely an inag ronnism. humans are not needed. the fast majority of stocks are traded here. to begin to understand the stock market and how it works, you need to go across the hudson river to new jersey. now the markets are almost completely computerized. networks and server farms at five sites are where most trades actually happen. i'm in new jersey, 30 miles from manhattan. the security building behind me is the new york stock exchange's data center. it's here, where almost all the trades take place. if you are a high speed trader and looking for a fraction of a second advantage, you have to be close to this as you can be.

that's called co-election. you pay to have your server in that building. onlile biggest players can do that -- only the biggest players can do it. the short distance matters, even if it's a few feet. having a connection that is a milly second faster to the competition could mean an extra 100 million for a high frequency trader. critics charge that co-location aspects is a form of front running. buying a stock before you do, and selling it to you for a higher price. it's drawing the scrutiny of state and federal legislators are concerned that it may provide advantages unavailable to investors. >> it's computerized trading depending on speed. in order to participate in that environment, you have to get direct feeds from the exchanges. one will cost you $60,000, just one. to get the participation passion

at that level, there's few that can. >> the need for speed does not end with co-election. high frequency traders pay for any microsecond advantage they can get. even one between two different exchanges. now, for example, there's a fibre optic cable linking the new york stock exchange servers in new jersey, and nasdaq servers in new jersey. a higher speed laser network linking the exchanges. >> the labour itself looks like wally, and accusation technology developed for the department of defense. >> when our network goes live. it will have the time it takes to move price information between the two locations. putting it in imperial terms, it takes 780 microsecond round trip time, and bring that to about 360 micro seconds. >> just to put that in

perspective. it's a 0.00036 of a second savings. imper sentible to you and me. potentially worth millions. this quest to trim a fraction of a second extends to the time it takes stock information to travel between london and new york stock exchanges. currently there are 12 fibre optic cables crossing the atlantic. the last engineered over a decade ago. the new route following a straighter line. >> we managed to shave off by building the cable. about 500km. which makes it about five milliseconds faster than the cable today. building the new trans-atlantic cable is not. the cost $200 million.

>> what is fascinating about the cable, the information is through the little fibre, of all the cable, the relevant part is the fibre pairs. others are to protect it. we curry power to amplify the signal, it's nas nating to think of the -- fascinating to think of the 5,000km. >> it doesn't end there. there's talk of constructing a laser network across the atlantic. using air ships. >> it's a success of three blimps or high altitude platforms making geostationary at 70,000 feet, talking to one another over the same laser technology, connecting new york to london. it will do it in 30 milliseconds the price tag for the trans-atlantic air bridge saving 30 milliseconds.

$500 million the machines and high frequency - meet the math geeks out of challenge building the black boxes, laughing out the way to the

well if you think a degree

in science engineering or maths is not safe. think again. the skills are considered total totally hot by wall street firms. they are mathematical formulas generating huge profits and controversy. the people that write them are quantitative analysts or quants. some colleges are training the next generation of traders by offering degrees in financial engineering. to better understand this phenomenon, go back to 1984, a colt movie called "revenge of the nerds." >> movie reel: nerds in "revenge of the nerds" maths and science students are pictured as dweeds, tormented by the jocks. back then the nerds under the power of computers that they could leverage to their benefit.

>> movie reel: what's your name that was 30 years ago. >> the matter of high frequency trading is a matter of timing. >> today the moths, science and physics students use their skills to join the ranks of the financial industry. at the institute of technology, they have a mag nif sent view and a campus that looked like it belonged in new england. >> don't be fooled by the setting. what goes on could help to shape many fortunes made across the hudson river on wall street. >> steven's institute of technology is one of about 100 colleges and university in the united states that offer degrees in financial engineering. the classes have titles like applied equations. at the heart of the programs is the ability to create high prequestionsy trading

algorithms, or black boxes, generating millions in profits for high frequency trading firms. >> they are very much in demand. they are capable of producing those. they may go on their own. much creating their own hedge funds. just ask one of the 97 masters students. >> we can design our own black boxes. having a successful black box is an atm in the sense that it's generating money. you don't have to be here right across the hudson river, new york university has a prestigious financial engineering programme, but not at the stern school of business, rather next door at the koran institute of mathematical science. >> this is a competitive programme toll get into. this year we got roughly 900

applications for a class size of 40 student. obviously it's more difficult to get into this programme than to get into harvard. >> the professor was in the strategy group at goldman saks, before coming to n.y.u. he says there's no substitute to beak in the financial capital of the united states. >> if you want to learn from the best, this is the place to be these students agree, and are willing to spend $50,000 tuition over 18 months. why? well maybe because graduates have gone on to work at credit swiss, morgan stanley, and a median based salary of $85,000, the first year out of school. more impressively, a few years down the line, with a little

experience under their belt, a quant might command a salary of a quarter of a million, with a bonus of $500,000 or more. >> rewards could be the size of achievement. you have done something great. and the work that comes with it. the financial strip, you know, if you did something great, it can be measured. i can disintegrate and do it with parts. >> with oo 100% -- a 100% placement rapt. it seems the opportunities on wall street are almost limitless for students with a degree in financial engineering, and it may be the in other words that have the last laugh. >> how fast are we going? >> the old cruise control is 35. >> that's our show for today, i'm ali velshi, thank you for joining us. saberi on a personal journey. >> this is the first time in 20 years i've been back to my mother's homeland. >> a special in-depth look at japan.

the legacy of the atomic bomb. controversial american military bases. and the country's evolving identity. behinds bars, a federal judge gaols a kentucky clerk who reverses to issue marriage licences for same-sex couples. >> because i'm not. >> under what authority. >> whose authority. >> under god's authority. >> reporter: saying she can't use her beliefs as an excuse format doing her job. human tashian crisis -- humanitarian crisis. gridlock and straining resources. e