Flash Crash 2010 – VPRO documentary – 2011


on May the 6th 2010 at 1400 hours 42 minutes and 44 seconds the US stock market’s go into freefall Dow Jones takes the fastest and most dramatic nosedive in its history an event that will be remembered as the flash crash where where are you on May 6 2010 uh here working May 6 2008 was I was here that’s been bull day literally locked away on holiday looking at the British general election 10:53 so 90 handles that’s too much that’s way too much I was on a conference call with a potential investor one on the market when I’m my own profits and losses and then my ears and mouths doing whatever else they have to do and the market basically looks like it’s going to go to hell that day it was the first really down day that we had in a while and I mark today let’s just stay on top of this cuz this could get really ugly so 40 minutes in we’re having in the middle of a great conversation and he says look I gotta go we’ll talk again this was really interesting but I have to go now and I watched the market fall another what two or three percent from there wasn’t just minutes later it was just Armageddon here you go I’ve told it to 835 the down ticker Wow almost a thousand points we call this a capitulate check uh I was just opening up my computer like most everybody and going oh my god I’m shouting to my colleagues cuz I have a giant to be in my office gawking at the screen like when the Dow hit nearly down a thousand points you’re like really hating if I didn’t Sharkey and I gave all that back and people didn’t know why it was happening and there was rumor after rumor after rumor and it kept on getting worse I saw that panic yeah I think then and then I started watching CNN or one of those what we’re seeing right now I mean it but maybe I believe maybe unprecedented I did talk about capitulation let’s take a look at P&G they’re gonna probably halt trading we can’t stop the selling I don’t think I found out about this in real-time so this was probably a half an hour later or an hour later there’s there’s nothing that one can do in the middle of the day so weak there’s no button that we press that says hey stop stop what’s going on so the markets are going to move the way that the markets move you just didn’t fit the facts that there was some big global news story it looked more like an error or a bad price that’s what it looked more like at first and then it got magically better and nobody had seen a recovery that quick either there’s a little bit like seeing someone hit by a car here you see some terrible motorcycle accident here the guy must totally be dead and the person gets up and walks away and about the market didn’t work it broke down the machines broke down that’s what happe posture was never 27 that to me is the mystery I think it really we we don’t understand these things this is really the earliest form of digital financial instrument George Dyson is a historian who has just completed a decade of research into the development of one of the first computers what I’m doing now is carving the data into this stick this notch represents 100 pounds of gold please go back to the 12th century and he’s smaller there’s a twenty pounds and ten pounds now breaks into two pieces that are unique this part stays with the exchequer with the with the Kings Treasury the person who brought the money keeps this part which is of course called the stock and that’s the origin of the word we still use of owning stock in something so if you put money in your bank or something you’re essentially getting a number that is uniquely matched to the money that’s in the bank and if you come back and say I want my money you get it it’s the same principle just just moving a whole lot faster on the morning of May the sixth the Sun rises in Asia and all is still quiet on the financial markets but the British elections and the continuing concerns about the Greek financial crisis are causing anxiety when the American exchange is open prices are under pressure at 2:45 p.m. many shares including those of Apple Accenture and Procter & Gamble collapse so you’ve got the general picture leave the big picture although it’s still over a very short time scale and then you’ve got the detailed picture of the individual trades a little bump there that’s almost the more natural scale even here at this level it’s something to be quite big so there are some trades that are kind of normal then there’s this yes fie these these these big right these are really big are they the way you explain this you can explain it Paul Wilmot is a mathematician and quant who is privy to the algorithms used by the automated systems driving the financial markets I think the thing that’s sort of sort of easier to explain is in the zoomed out scale and you can see this nice falling off a cliff and then bouncing back up yet the thing what’s harder is these these spikes that the trades and the trade happening down here but every now and then there’s something up there which is sunny and the difference with our new mallocing enormous differences but what’s causing it and I don’t know what’s causing it this one go it starts off at 42 and it goes to all the way down to one what sense Lord above it’s amazing it is better to fail conventionally than to succeed unconventional Wilmot gives classes to traders hedge fund managers and bankers for years he has warned about the dangers of an algorithm induced collapse at the financial markets I would look gosh I turned away can that happened so who is it selling extension 5 cents actually there is rocketing back and forth miss me yeah only a machine can do this right any thoughts of fun what’s going on here Steve Jobs didn’t die there’s no reason for a before falling to zero right and so they start trading and then I mean the back up in what seven and a half minutes or something it was incredible and that what couldn’t many people not reacting that quickly and not in that kind of volume well one key thing is how automated is all of this all these trades can you identify that this was caused by some some black box thing or was it somebody in the old-fashioned way picking up a phone etc I’m making a trade I go in my head a scenario whether there are no but there are no people at all anymore there are just there are black boxes and they’re really clever buy black boxes and designed by famous banks that we won’t name and not too clever back black boxes and one day the clever black boxes decide to stop trading because they don’t like the numbers are coming in leaving you with a quite dumb black boxes they’ll do basically anything accents you should not be selling I have a couple of cents I mean it just doesn’t make any sense that’s all so if you just left with a bunch of dumb black boxes they could be doing anything you know the Vatican selecting the Pope is maybe a pretty black box kind of a process they take in inputs about the candidates they God knows what they do in there and then there’s some chimney smoke and boom it’s done right there’s your output Rishi Narang manages a fund investing in a global network of high-frequency traders he is a strong believer in automated financial markets so it’s any process whether it’s systemic or not that that has inputs and an output but where the process of transforming the first into the second adjust is unknown or unknowable so till the black box is a box I mean these modern black boxes are boxes that sit at the at the stock exchange and has some rules or algorithm I mean it’s and it’s concealed within the box that decides on Wednesday I’m going to buy this stock and on Thursday I’m going to sell it so we don’t know what’s in these black boxes and that’s the that’s really the interesting thing that you’re like it’s like a potential poker game where you don’t you can’t see the opponent’s cards but you know how much they’re being so this is a this is a big financial service room here so no name on the door nothing you would never know what’s going on here as you notice it’s guarded gated there’s cameras everywhere they probably know that we’re here right now so in a period of time though they would come out and say what are you doing stop doing that type of thing and if you rolled up and said I’d like to come in they would not let you in jeff hips men once employed by Wall Street investment bank Lehman Brothers now works as a data center broker for the world’s leading financial companies and obviously and and also no signage so you don’t know whose facility this is you don’t know what’s going on inside of it you know completely a nondescript building no windows no windows if this building has a problem that this building is damage to this building if anything happens to this building they run the risk of not being able to execute trades so they protect this this is the heart and lifeblood of their business so imagine your home where you have double everything double power utility feed coming in two refrigerators two washing machines two televisions and if something were to break you can go immediately to the second second device in the data center world that’s called a Tier four facility and what they need is two of everything because the cost of downtime to these organizations if they’re not able to trade or they lose trading operations for seconds it could cause some enormous amount of money some of these firms take it to the next level where not only do they have it with a duplicate within a facility then they have a second facility that’s exactly the same the goal of these facilities is to minimize if not eliminate any single point of failure New Jersey is a very hot market for data centers because of the distance issue to Manhattan when you’re dealing in milliseconds there can be absolutely no delay or being that the technology has a limitation in terms of how far away can be in terms of that latency it’s based on the speed of light between one point another that distance becomes your outside ring or outside distance that you can go tends to be in the neighborhood of thirty to thirty five fiber miles so we’re constantly tracking the fiber as to where it is and what the main routes are and we also track with the power is because another main component of these facilities is they are very heavy power users they need an enormous amount of electrical capacity in order to operate we don’t want to have any outside event caused an outage in the facility this may be things like highway access can a tractor-trailer accident caused an evacuation of the area look in flight patterns we look at flood plains we look at nuclear plants we look at rail lines we’ll look at crime levels we’ve looked at lightening strikes the goal is to eliminate any possibility that one event can cause an outage in the facility so typical facility could be anywhere from three to four hundred million dollars you know the wealth can be distributed evenly or wealth can be distributed unevenly and wealth can be distributed unevenly among people and it can likewise be distributed unevenly and evenly between people and machines I mean all the money is flowing to the computers and the same with these stock exchange facilities in New Jersey them the money is flowing through these points and the machines are surrounding it you’re just sort of siphoning away that that wealth but if you go into the neighboring communities the people are really very poor high-frequency traders are estimated to generate more than half of the entire equity trade volume in the US traders or rather computers making money by buying and selling within a matter of milliseconds if you think about the market and the center of the market being the innermost circle the specialists or the market makers at the exchange the exchange members and then there are these concentric circles as you go further and further out and the last circle is a mom-and-pop investor that’s not in a financial center they’re in the middle of a farming country or something or farmland the speed that the highest frequency trader knows something and can act on it before a discretionary trader sitting on his farm as home PC is now a matter of a second or two the set also mean that that high-frequency traders has to earn his or her money within these two section yeah it does yeah because by then the rest of the world is processing the same kind of information they call it the race to zero because zero is the amount of latency or delay so if you think of this in a really metaphysical sense in the actual market some an event happens someone raises their bid from a hundred dollars to a hundred dollars in one penny how fast can you actually receive that information that’s the first question the second is how fast can you process it and the third is how fast can you get your your decision back to the marketplace to act on right so this is why speed matters I’m interested in the origins of the digital universe because it’s this world that affects us all now we all we live in this digital world and how I mean interest in the beginnings of things how did it start so this particular computer that I’ve spent so much time interested in is not the first computer but what’s important was it was the first randomly accessible electronic memory that’s the critical thing the size of this matrix was 32 by 32 by 40 bits that’s five kilobytes that’s now the memory we use to display an icon on our desktop and I think in an mp3 it would be half a second of mp3s that was the size of this universe they created up until then numbers represented things say well I have ten oranges or ten apples or the temperature is 10 degrees now you had numbers that were order codes that could actually execute instructions in the machine so numbers were allowed to do things these are the crew of the private guys who and women who really made it work the engineers machine shop people the women who did the coding and their name for it was maniac mathematical and numerical integrator and computer on the maniac they did the calculations that led to the first hydrogen bomb test which was called was called IV Mike you can’t just build a small hydrogen bomb and see if it works and then build another way it has to work the first time and nobody could and I think they could have done it by hand but it would have taken 20 years that’s what’s I think so profound about this financial thing because suddenly it used to be well this is sort of interest to people you know designing nuclear weapons or studying the spectrum of light coming from stars only who really cares about what happens the frequencies of a billionth of a second it just it really only concerns physicists now suddenly it you know you it concerns all of us because it’s it’s gone into this financial world that represents all are no or wealth I only see five hundred seventy why isn’t it updating alright alright we’re in Belmar debt on May the 6th 2010 at 1400 hours 42 minutes 44 seconds and 75 milliseconds an email a future traded on the Chicago market begins to show erratic price swings the fluctuations affecting this emini an important indicator of market moods soon spread to other shares within the US ultimately leading to the fastest and most dramatic fall of Dow Jones ever nearly eight hundred and sixty two billion dollars go up in smoke albeit briefly there were two things that stood out a lot one was the speed that the market initially plummeted Gregg Berman a Wall Street veteran himself led the investigation into the flash crash from SEC the US financial market regulator so for broad indices to plummet multiple percent three four five percent in a matter of a minute or two five minutes that in itself is quite extraordinary the second aspect that was even more extraordinary was the number of stocks that were affected and the depth at some of these fell – so when you see large cap stocks fall 60 70 80 90 % that was quite dramatic at that time what we didn’t understand and again we’re talking about just two to three hours after the event is how can stop straight for a penny how can stocks trade for 1/100 of a penny who’s buying and who’s selling down there so when this emini drops then the SPI drops well the s py is made up of 500 different stocks and then all the stocks in there are members of indexes different indexes and so though we price and then all the whole mass has options on top of it and for s py is pretty much all the options so those all have to be price so you try to repress all this stuff in 50 100 milliseconds and the system cannot handle that evanston a city just north of Chicago is home to Eric Scott hyung Saders company Han seda is a data analyst whose company collects purges and sells financial data on May the 6th he received urgent reports from clients complaining about incorrect share price information the markets could no longer sustain the massive data flow produced by this flash crash and that’s when the delay started a few hundred milliseconds and then a few seconds and then 5 seconds and then 10 seconds some of them went all the way into 36 seconds so the delay into that of a 36 seconds right but the way that the timestamps occur is it hot hit the delay so you couldn’t see it the Dow Jones is actually should have been here the real price was here the delay in the system caused it to appear like it was over here so when people were seeing the Dow dropping this was actually really where the market was at that moment in time what we’re seeing right now I mean it may be I believe may be unprecedented or dad to talk about capitulation let’s take a look at P&G alright this is gonna say everything P&G is now down 25 bro that’s should stock is there you just go and buy it daddy’ll be there to watch people talk about a major business news channel who have seen mark is day in and day out and they didn’t know that this market already turned and was flying higher something’s not right you should at least be able to know that you’re behind it’s one thing to be behind it’s another thing to have no idea your pod long before the SEC report came out Eric on sader had published his first findings on his website we’re the ones who pointed out nice he was delayed on May 6th and they testified that they weren’t a couple times so but you know the SEC report said it didn’t matter so I guess didn’t matter but if your clients are accessor matters to everybody I know that raised that’s you know the name of the game is getting getting the data faster and they require everybody who connects to them to have a thousand-foot cable no matter if they’re 10 feet away or 50 feet wet because they all would fight over who was closer so a thousand foot cable it takes the speed of light one microsecond to go down a thousand foot cable so if somebody had a 500 foot cable they would get one half of one millionth of a second faster but latency doesn’t matter does it only those financial institutions that use the much more expensive direct data link to the New York or other stock exchanges received the price information without the delay that could run up to 36 seconds this is the NASDAQ exchange see the exhaust from generators and then the louvers or the air intake so they’ve got three in place and looks like they have a slot for another one so it’s being constructed right now right right you see security they’ll have gates around the whole thing the generators are in the back and then this is another financial service firm no windows yep we can go thro till they yell at us is their cooling systems right back here you can see the generator exhausts in the back and then you’ve got security gates don’t get these guys freaked out when they see a camera and is there a reason that they’re so close to domestic no they were in first Nasdaq came in later it is possible for hedge funds and banks to make money from all this volatility that we’ve got in here it’s let me stop supposing somebody bought here and sold here they just made a fortune and they do that they could retire on that accenture when it’s down at cents a few cents when normally it’s up at $42 I don’t know who was doing this it would be nice to know maybe not a who but or was indeed yes yeah if you had GE I’d be really interesting because they bid be higher than the end no you see that happening girls oh yeah right side and phone cuz they got behind five seconds at leche but what does it what do I see you verily you’re seeing what exchange is there there big price is higher than the others offer prices so you could buy it on the lower one son on a higher one immediately and capture that profit you don’t want that tap it you don’t want it to be the case that the S&P is a thousand points here in a thousand and twenty points there it’s the same index it should just be a thousand okay okay so let’s say I can buy the SP here for 10% down on the day but I can sort of sell it here at the exact same moment for 8% down on the day I’ve made two percent with zero risk if you have the exact same instrument price differently in two places that’s free money this opportunity actually existed so people could and they were buy one is on the other so that’s making money of the malfunctioning of the system yes yes it’s arbitrage ight once again seventy-five even off beyond our guys revenue even awful guy he’s probably he waited he has ideas down whoever even though it was a five-minute flash crash it took five months for us to do the full investigation the first is we had to organize so we developed and deployed a multi-divisional cross-functional team a big team consisting of people basically from every division at the SEC we also coordinated with the CFTC because this was a joint investigation and they had a multidisciplinary team on their side as well we had to basically reconstruct and understand what happened what’s the order that events happened in and hopefully from that we’d be able to determine why it happened while the SEC was assembling a multidisciplinary team Eric ham seder continued his own research into the market delays so what are we looking at here I wanted to show how an order that comes into an exchange of that price gets transmitted to all the other exchanges so I built this model and represent each exchange by one of these boxes and the price is in each one of these exchanges shows how that exchange sees the bids and offers on all the other exchanges at any given moment this is just for one stock there’s a lot of information that occurs in the marketplace effectively were to take the first second of trading for all equities options futures the United States and you printed it on a sheet of paper so each line was a trade or a quote the first second of data would be a stack of paper about six feet high one second yeah ones one second would be about six feet high it’s boggles the mind like on that day was we had seven point six billion data points you know billion is a big number so we take data from all the exchanges we save it archive it so that we can replay any day exactly as occurred whenever we want it would look to us like nobody was fighting what was going on it said you know what we are in a unique position to actually probably figure this out because we can play this you know but you just throw everything at it and see what’s so unusual and so that was the thing that sucked us in really five months after the crash the SEC report was published identifying the cause of the flash crash as an uncommon ly large order from an investment fund for the immediate selling of 75,000 emini contracts with a total value of 4.1 billion dollars long before this report was published it had leaked out that the source was the Waddell and Reed investment fund based in Kansas City where we know from wall street we are a thousand miles and we are culturally a million miles from all street bill black a financial regulator back in the 1990s now works as a professor of economics and law at the University of Missouri in Kansas City Kansas City became important as a matter of finance because it became important as a matter of Agriculture and as a matter of foodstuffs especially meat it was the place where you came from Texas what in all those western movies with your cattle because the railroad was here and because the mighty Missouri River was here and that’s what made Commerce so easy railroads had all this right-of-way in nice straight lines where you could lay the very sophisticated cable that allows you to transmit enormous amounts of information very quickly and very cheaply and very reliably the question remains whether it was Weddell and read that by using these cables to place its major selling order cause the u.s. stock markets to plummet are we at Ground Zero is this the place that triggered the flash crash well the Securities Exchange Commission in the United States and many other people say yes this is where it all happened and that it was triggered by some trades roughly four billion dollars by Waddell and Reed in a particular index one of the things that I think people find hard to believe is how can any order move a market a lot of the trading that goes on is a market neutral I’m trading with you your tray of me back and forth back forth tremendous amount of volume but neither of us hold more than a small amount at any given time I show a little I buy low I sell I pilot I sell low I buy a little and then someone comes along and they say I sell a little I sell a little I sell I sell oh so wait so that that’s that completely changes the dynamic where’s that where’s the buy part the behavior of all of the market participants changes because of this very very large trade what Alan Reed says it wasn’t us and and what can Waddell and Reed do right the more they talk about it all they do is create publicity and so they haven’t followed that strategy yeah hello you’re on the security well done Reid yeah what else very they don’t like to talk much but internally you know as humans are some of them were much closer to it or like I can’t take this anymore and so they had seen the analysis that we had done and they they said would you like to look at our data and I was like yes send it over I didn’t really believe it was could come over like that and sure enough it did and then I lined them up and saw how they was executed that was the shock for me because then I could you know I could cut through all the the talk and to see the data because the data will always tell you what happened if you just look at the data there is always room for multiple interpretations this is not a flash crash issue this is you take any data set anywhere and you can have five people and you’ll have seven different opinions this definitely resulted from from that large trade and the way that that large trade was placed you can’t see that by looking at just the information that that I think other folks have looked at rode some software that took all the trades in that period of time and I plotted it out to see where it traded and I looked and I was like so you see the red dots here that’s the Weddell and we these are were word would on we traded here so it’s all the blue dots are where they did not you and Sam on that one you’re not I’m in there I don’t nest out there there’s one little dye here and there’s one here one here again only when the markets going up they’d only sell when the market goes up they don’t follow it or chase it down so this algorithm actually is very sophisticated it’s extremely sophisticated I mean it would never sell it the bid price for example it will only sell when somebody was willing to pay that price buy it you just a couple here here they’re selling it on the way up and then it stops here nothing in this case what we were seeing is that the patterns were such that things were happening at the second and the one minute level so we’re able to take a lot of the quotes that were happening at within a millisecond and aggregate them so the data that you actually see in the report tends to be chunked up into one minute blocks so you can look at it the millisecond level but the the noise that you get in that millisecond time frame doesn’t tell you much this is on a 100 millisecond basis because if you if you go out to go out to a second for example then this will be averaged over the whole second so it’ll dilute about ten to one and if you looked at it in a minute well forget it you would never see this so he just wouldn’t show up but does that mean to you that the SEC report is nonsense in that sense well here’s the theory here’s here’s what they missed a kiss and this is this is when we this is when it came out this is when we kind of had an epiphany it says that the algorithm sold to buyers and some of those buyers were and I don’t want to get myself in trouble the whole group of people so let’s just say buyers one of the but one group of buyers would accumulate up to about 3000 contracts and when they got that that limit then they would turn around and they would sell it sell that position and what the report doesn’t mention is when they sold it they didn’t salt like the Waddell Reid algorithm did which would wait they sold it as fast and as furious as possible this boom down this was instantaneous like in 1500 milliseconds they will sold about 125 to 150 million worth of these contracts having here and having here and then and here and here and here and then BOOM and so my question is if you’re gonna name Waddell and Reed or not name a but name them why are you talk about this but to me that is those are those are very disruptive in fact they’re so disruptive there’s the that is what caused it so this thing started that’s exactly when the market delays happened and they just snowballed from there so it’s one firm I don’t know no they would know you know I would would know the SEC they would know because for them the data are they have they have okay so all I get is time price a sequence number in size they get that plus they get who did it there are some very strict disclosure rules about what can be disclosed and what cannot be disclosed and we are subject to all those disclosure rules both at the SEC and the CFTC so you are not allowed to mention the names that’s a level of detail that we were able to go into was the level of detail we were we were legally allowed to go into or didn’t want to protect somebody I don’t know I mean I’ve done all you do we took that first step the first step is that okay let’s commit a week or two to this let’s just you know get into this it was being able to commit that time you learn just enough that Oh God you know it’s just so close we got it we got a look at this and then it just it really was been like that all the way I still see the light again to total but you know I almost regret that you started wondering well I’m thinking it’s a train coming at me six years ago there were many trains a day going south from Canada completely full of lumber you know to California to built to build all the houses that now can’t be sold and now you don’t see any lumber going that way coal going north to to Vancouver and then to China we like to think because they’re not in our lifetimes this digital universe has been running faster and faster and faster every time you buy a computer it goes twice as fast but but but in the digital universe there really is no time the time time as we know it just just does not exist in that universe computer is not operating on time it just operates on sequence something happens and the next thing happens the next thing happens you ask for the next instruction execute that instruction go to the next one and now we imposed it’s sort of like circuit breakers we when we build our micro processors we say ok and you can’t do anything until the clock says it’s you know a nanosecond later and you could do the next step but the computer actually don’t need that at all they they could be be totally asynchronous and just you know so we sort of impose our clock on their system what is the implication of that well the implication is that that this other world exists now it’s not not tied to our form of time at all what we had found is when markets move fast a lot of market participants pull back and while that sounds like a sort of a simple thing it’s that’s pretty much it when market participants hope they pull back when markets move fast so how do you solve that well how about saying when markets move fast we’re going to stop the market from trading to give participants a time to come back in first we spoke with all the exchanges and we met with a bunch of different market participants and very quickly you put in what’s now been known as the circuit breakers which for certain sets of stocks and for funds we pause the market for for five minutes if the price moves more than ten percent in five minutes the SCC’s circuit breakers though these five minutes is like a neon and a long time and you know they sit sets it’s setup the people could actually game them and cause delays in certain stocks by just simply forcing the price to go high enough it would be they would do so only if they found that there was a way of capitalizing on that if you’re the one who causes it and you know anytime you know what’s going to happen for somebody else does I mean that’s just opportunity and so they’ll be at the end up being magnets so when you get like if it’s 10% breaker and you’re at 9% you’re most likely going to see that 1% close a lot sooner than it would otherwise we operate on this time scale where you know we’re five minutes is pretty fast you could go out for a cup of coffee come back and find that you’d lost a billion dollars that’s that’s that would ruin you know ruin your day and yet for these computers operating in microseconds it makes it makes no difference so I think it’s the circuit breakers is the case of sort of in imposing our time scale on the computing time scale could it happen again the same mechanism that caused it on May 6 is still there in fact we see it actually occurring more often where there’s a at least weekly there’s a pretty strong move in the e-mini is symbol very similar to the first event we just don’t see them back-to-back coupled with a lot of bad news and we can’t we really can’t predict we don’t know what we don’t know inference is how the algorithms have changed now now that you know everyone have may have rewritten their codes to take advantage of the next time this happens so the next time it happens it will be different there must be a control room somewhere overseeing everything that’s happening on the financial markets somebody no I don’t believe there’s any Wizard of Oz behind the curtain no there is no air traffic control for the market and there is there’s no one person or one group that in real time sits and watches and says wait wait that that water over there hold on let this order go through first it’s so complicated just the volume of data alone it’s just so overwhelming that and the problem is is that it keeps growing at this hue it is a alarming rate the technology that would be required in the skill level that would be required to monitor markets across exchanges in multiple places at super high frequencies and know how to deal with it the technology and know-how is typically not found in government’s because it costs a lot and taxpayers probably don’t want their money spent that way so you won’t be you won’t be able to find people who would be willing to do that work for the for free effectively not so much I don’t think so the crash I expected but not the recovery it was the recovery that was a surprising thing how did this thing suddenly go back to a more or less stable state it is a little scary that that that we’ve given given these machines so much power over our lives on the other hand you can take the other view that machine management of these systems may work a lot better I don’t mean that’s the optimistic look humans are great because we invent things like computer trading strategies and auto pilots but we’re not so good at executing a thing over and over which is why we have machines build our cars once we design them it’s why we have computers fly our planes once we design them and design that algorithm and so on it’s just the same thing applied to markets maybe they’ve watched the Terminator movies too many times I’m not sure but as soon as anything sort of slightly odd happens in the markets it’s immediately let’s blame the quants let’s blame the computers computers they do what you tell them so if you understand what’s been told to them you understand the strategy so who knows what’s inside the black boxes some you hope somebody knows but we don’t even know it’s quite it’s entirely possible and I’m sure that there are companies doing it that are allowing those algorithms just to evolve on their own just letting the black box try different things it with small amounts of money and if it works reinforce those rules it’s that’s that’s been done we know that’s been done so then you actually have rules where nobody knows what the rules are because the algorithms create new rules for themselves yeah you let them evolve the same way nature evolves organisms need to say that well every one out of a thousand times you change something to see if it works better and it’s amazing how quickly you can you can evolve algorithms that work better if you are an investor meaning that you wake up one morning when you say I would really like to buy a stock that I’ve been tracking and I really understand that very very well then you can buy that stock owning the stock market is historically from the data I’ve looked at and studied not been a particularly clever idea for a really long time the last time we traded was I bought Apple in 2001 I’m not afraid because all my money is in there in cash and no interest I don’t have a dollar in the stock market but it interests me just as it you know so I’m sort of happy when it goes down and you’re tasting when it goes up why are these other people making you know making money by not doing any work stocks in the u.s. just in the last 10 years have had two distinct declines of north of 50 percent so to make 5 percent a year you’re risking half your money that to me is an awful trade I don’t make it you

You May Also Like

About the Author: Oren Garnes

Leave a Reply

Your email address will not be published. Required fields are marked *