Apologies about lack of blogging, but my job is taking a heady chunk out of my life, and when I get home I am simply too tired to do anything but plop down and doze off.
This blog is dying, it is on life support, and is very much down from the heady days of 300-500 hits per day - I am truly sorry about that as this is my baby, and I am proud of it. I went back to reread some of my older articles and whoa! they are good (am unashamed to say so).
Blame my work on the lack of articles.
The over an hour one way driving in traffic does not help.
Europeans work to live, Americans live to work - truer truism was never spoken.
The financial "experts", they are everywhere in our world.
On radio, in newspapers, magazines, books, and of course, the idiot picture box.
They do not hesitate to tell you what to do with your money, whether to go with bonds, stocks, what industries to get into, and much other advice.
But all of them, whether TV anchors coke snorting, manager's cock sucking bimbettes from CNBC, CNN, ABC or the "real" financial "gurus" who specialize in peddling misinformation, like Jim Kramer (do me a favor and click here, it's a doozy - one of my oldies, when I was a decent blogger) or Suze Orman...
They all begin their programs, or news segments, with the same routine.
Today the markets moved up or down, gained or lost, based on a following tidbit...
According to these experts, the stock market in America moves are based on specific, tangible items in the news - whether a job report released by the government, or a lawsuit against a company, or...
Well, pick whatever reason - it changes day to day.
Oil rises on economic data, market gains based on the upbeat dollar, market gained because PetSmart beat estimates, etc etc etc.
The reasons for markets moving up and down are endless, varied, and very creative.
Every day something happens to move the market ticker up or down.
But sometimes the reason given is much too small to sway the markets one way or the other. Granted, an unemployment/jobs report released by the government is big and strategic, but many other reasons (PetSmart beat expectations!) do seem to smell a little bullshitty...
It's almost as if the "experts" have to find a reason to every day for the market activities to explain why market gained or lost, why some companies' stocks went up and others down.
Like they are following a script.
There is one thing that every "expert" in our wonderful mass media is not sharing with their audience, an important thing that affects the market in a major way, every day.
The fact is that from 30% up to 70%+ of all trades on the stock markets are done by algos, computers running algorythmic programs/routines.
Now, I am not talking about Joe Schmoe going online and day trading.
I am talking about big hedge funds, pension funds, big investors like Soros or Buffet, using software to front run others.
Wikipedia has a good overview of the phenomena.
Lets dig in.
A third of all EU and US stock trades in 2006 were driven by automatic programs, or algorithms, according to Boston-based consulting firm Aite Group LLC. As of 2009, high frequency trading firms account for 73% of all US equity trading volume.
The naysayers will say that this is Wikipedia, it is bullshit and is not to be trusted.
The New York Times, March 5 2010:
Computerized trading of stocks first became a significant part of the Wall Street scene in the 1980s, when it was lamed for exacerbating the market plunges in October 1987. Since then, the computers involved have grown vastly more powerful and the algorithms that guide their trading vastly more sophisticated. Automated trading has also come to play a far larger role in the financial markets.
An explosion of computerized trading has helped drive volume on the New York Stock Exchange alone up by 164 percent since 2005. Stock exchanges say that more than half of all trades are now executed by just a handful of high-frequency traders, who use rapid-fire computers to essentially force slower investors to give up profits, then disappear before anyone knows what happened. High-frequency traders generated about $21 billion in profits in 2008, the Tabb Group, a research firm, estimates.
Using the allmighty google re: 'algorithmic trading' does not bring answers, it brings questions and more questions.
The main one is "Exactly what percentage of all trades in the stock markets around the world are done automatically by computer programs?".
The answer is that no one knows for sure, but at the least 30%, with the USofA reaching as high as 70%+.
Lets see what the government says.
The answer is, even the FED does not know exactly.
The Federal Reserve publication, Last update: November 3, 2009:
Aptly entitled 'Rise of the Machines: Algorithmic Trading in the Foreign Exchange Market'...
The study about algos trading was done for the years 2006 through 2007.
First, we find evidence that algorithmic trades tend to be correlated, suggesting that the algorithmic strategies used in the market are not as diverse as those used by non-algorithmic traders.
That means that the computer trading programs work in similar ways, and have similar ways of operating.
Since all these dogs are chasing the same rabbit, speed matters, as winner takes the greatest skim off the top off the trading cream.
That is why the trading desks in our banks (sorry, "banks" - they are not lending anymore to businesses and citizens, they are shoveling money into their Excel spreadsheets that feed algo trading), hedge funds and pensions obsess about "latency" - the speed of the trade.
Think on that - all the algos for a specific market segment (oil, or currency trading) work in pretty much the same way, and the winner is the fastest one in the pack.
Does that make for a healthy trading environment?
In algorithmic trading (AT), computers directly interface with trading platforms, placing orders without immediate human intervention. The computers observe market data and possibly other information at very high frequency, and, based on a built-in algorithm, send back trading instructions, often within milliseconds.
Consider: you, Joe Schmoe the idiot feeding the mammon beast stock market off your dinky PC computer, are front run by every single one of these computer algo trading programs, who front run each and every one of your trades to skim a bit off the top, eating up into your profits.
That is what you are up against.
Among the most recent developments in algorithmic trading, some algorithms now automatically read and interpret economic data releases, generating trading orders before economists have begun to read the first line.
This article was written in October 2009, by the by.
Terminator, we are here.
Financial Times, February 18 2010:
Not long after lunchtime one day on the New York Stock Exchange three years ago, unusual things started to happen. Hundreds of thousands of "buy" and "sell" messages began flooding in, signalling for orders to be made and simultaneously cancelled.
The volume of messages sent in was so large that the traffic coming into the NYSE from thousands of other trading firms slowed, acting as a drag on the trading of 975 shares on the board.
The case was made public only last month when the disciplinary board of the NYSE fined Credit Suisse for failing adequately to supervise an "algorithm" developed and run by its proprietary trading arm - the desk that trades using the bank's own money rather than clients' funds.
I am 110% sure that this was explained away on the idiot box at the time as some inane fact, perhaps even by PetSmart beating expectations or some nonsense.
A decision to trade can be triggered by a news event; there is even separate technology that "scrubs" news articles to give algorithms a sense of where - "directionally", in the jargon - it may be profitable to start trading in a company's shares. Other types of algorithm seek out where, across a range of exchanges and trading platforms, the best price may be found.
Again, these things are hooked up to google, and are programmed to search for words - and combinations - and start trading on their analysis.
IF General Motors stock > -0.25,
Wait 0.000005 seconds
Wait 0.000001 seconds
Imagine now if an article made a mistake (these are written by humans, after all) and a minute or a few minutes later a redaction comes in on the web "Sorry folks, we made a mistake, the article should read...".
Meanwhile the algos went onto their warpath against each other and, more importantly, against the stupid, solitary Joe Schmoe idiot investors to front run them based on this news.
(The worst and slowest front runners are the brokerage services, online and off, who also, ALL OF THEM, front run their clients, every trade that they can. So not only are you being beat to the punch by the computers, but also by the relatively slow humans running your brokerage or online trading website).
The speed at which such trading takes place is causing alarm. The technology is so sophisticated that thousands of orders can be sent to an exchange's "matching engine", where orders to buy and sell are put together, and a match found, all in less than 300 microseconds - 1,000 times faster than the blink of a human eye.
There is absolutely no risk there, as we all know that human programmers never make a mistake, and even if they do it is caught by the Quality Control Team.
Why, you will never find the fact that I personally shut down a production line for over 5 minutes due to some wacky code that slipped through (Wait a minute! Forget what you just read, us programmers are perfect - just like Algo trading computers).
At the same time, markets have come to be dominated by "highfrequency traders" who rely on the perfect marriage of technology and speed. They use algorithms to trade at ultra-fast speeds, seeking to profit from fleeting opportunities presented by minute price changes in markets. According to Tabb Group, a consultancy, algorithmic and high-frequency trading accounts for more than 60 per cent of activity in US equity markets.
60% of the market is run by these computer programs.
Again, no one knows for sure, and we are guessing on perhaps one of the most important developments in our world today.
And of course, with that kind of money on the line, errors are surely caught uh oh...
In his letter, he identified a number of errors, such as Morgan Stanley submitting a $10.8bn order instead of a $10.8m order in September 2004 and a $31bn order placed by UBS in February 2009, which was 100,000 times larger than intended.
Looks like Steve in his cube missed a few zeros in his code...
It is a worldwide phenomena, as the human element is rapidly being replaced by automated scripts running on computers:
TheAustralian, October 14, 2009:
The Australian Securities and Investments Commission last month told the Australian Securities Exchange to investigate whether so-called algorithmic trading was affecting the orderliness of the market and whether tighter controls were required.
It is estimated that 26 per cent of buy-side traders use algorithmic trading for more than 40 per cent of their order flow.
One of the problems of algo trading is that it further concentrates power in the hands of the ultra powerful few, with the rest of us schmuck traders forced to settle for the scraps.
Another problem is that this does not help the market transparency.
Let me explain.
The theory of capitalism and free market (what a misnomer) is that trading helps us arrive at a price for a good or stock that makes sense, and it is arrived at because everybody can see how a stock is trading, who trades what, who buys, who sells, how much is in motion.
That is, perhaps, the reason for the existance of the stock market, and any market anywhere on this planet.
With algo trading, with hundreds of thousands of shares being bought and sold in less than a second, good luck on that price discovery thing.
Here is another perspective (written it must be said by a disgruntled trader);
High frequency algorithmic systems have been programmed to step inside the NBBO (National Best Bid and Offer), and be the best bid and best offer. This puts the computer system at the front of the line to be first for execution, and gives the computer the best chance to capture the spread. Unfortunately, this practice is dominated by a few large firms, and they have driven traditional market makers out of the market. If a traditional market maker places a bid, the computer automatically steps in front. In some cases, it steps in front by as little as 1/100th of a penny (a practice called sub-pennying, which is discussed on my website http://www.defendtrading.com).
These programs are very predatory and step in front of the NBBO on a constant basis. This has driven liquidity providers out of the market. Our proprietary trading firm, Bright Trading LLC, in the early 2000s, used to account for 2% of the volume on the NYSE. Now we account for just a fraction of that. Our 400 traders used to provide a substantial amount of liquidity to the market. But due to predatory HFT market making practices, we now provide very little liquidity. We are now liquidity takers. The rational is simple, if we place a passive limit order (providing liquidity), the HFT algorithmic programs simply step in front of us. If we do get filled on a passive order, it is almost always because we are wrong. You are sub-pennied when you’re right, filled when you’re wrong. Hence, there is no point to us providing liquidity. Other proprietary trading firms, floor traders, and specialists are in the same boat. There is no way for them to compete with the algorithmic programs, so they don’t place passive orders.
The Rambus incident:
Rambus (RMBS) fell 30% today in a matter of five minutes. It immediately bounced back. The cause for this move was speculated as a trader with a “fat finger”. A trader simply messed up and sold too much stock accidentally, causing a swift and violent sell-off.
Without traditional market makers, willing to step up and be the buyer of last resort, we risk having more incidents, like the Rambus incident.
Computerized algorithmic market making works in any type of oscillating market, as the computer can keep flipping out of it’s longs, and covering it’s shorts. It works in a trending market, as long as there is some type of choppy trade. The problem lies, when the computer system can’t flip out of the position. Most algorithmic systems are programmed with some type of risk parameter. If this risk parameter is breached, the computer will dump it’s position and cut it’s losses. This is what may have happened in RMBS today. An algorithmic system making markets on the long side, got too long, and was unable to wiggle out of the position because of the follow-through in selling pressure. Once it was down so much in the position (the risk parameter was breached), it dumped. This simply added fuel to the fire. That is why the sudden plunge to $16 happened. If you check the chart, you will not see this, because Nasdaq busted all trades under $22. But don’t kid yourself, these trades happened, and we should be very alarmed, because it will happen again, and it may happen to the entire stock market.
I particulary "dig" the fact that the stock exchange busted the trades - oops, it was a mistake, lets back up the trades folks!
Boy, if only my life could be this way.
Say, I buy a car, sign the papework at the dealership, get the keys, happily get in, turn the engine on, and drive back from the parking spot... cruuuuuunch, just hit a metal post, with the fender falling off.
Fortunately, I just back out of my trade, I didn't really just buy this car (heh heh), just like the stock exchange does - oopsie!
Business Insider: How The Rambus "Fat Finger" Incident Was Really An Algorithm Gone Wild:
Earlier today we had the chance to speak with Ben Bittrolff of Cyborg Trading and had the opportunity to discuss the world of high-frequency algorithmic trading at length.
Eventually we came across the topic of the Rambus (RMBS) "fat finger" incident in which RMBS stock fell nearly 35% in a matter of minutes, only to later rebound.
Traders called the incident a "fat finger" or human error. According to Ben, such is not the case.
As Ben argues, human interaction is largely gone from major trades involving large blocks of stock. An algorithm will not allow a huge order to go out all at once. If anything, it'll slowly begin to breakup the block into smaller orders, whether you like it or not. The algorithms have fail-safes of sorts built into them, so when a large caliber order comes through, it's immediately checked and processed into smaller orders.
A better theory for Rambus is that incorrect data was fed to an algo, setting off a chain reaction across the market. All an algo needs to go haywire is a false tip or data point and everything goes off.
All I gotta say is that Reuters and Google better double check their articles, because if, say, some hacker or prankster slipped into the system and changed the first AP article (the progenitor, the one that everybody else, from google on down, pastes and copies into their editions) and changed, say, the government report that stated that 36,000 jobs were lost in february to, just for shits and giggles, 360000000000000000000 jobs lost...
Well, it would be fun, wouldn't it?
I am sure that the "experts" would explain the market action that day as being caused by something inane.
PetSmart expanding into the Chinese restaurant market, perhaps?
Bonus: More Algos Gone Wild!, courtesy of Zerohedge blog.
Credit Suisse HFT Algo Gone Wild Slapped With Whopping $150,000 Fee
On November 14, 2007, the SmartWB examined over 800 securities and identified 129 securities that it wanted to trade based upon the algorithm’s parameters. At 3:40:01 p.m., the SmartWB began trading and immediately began receiving executions. After approximately 45 seconds of trading, the Firm trader/programmer changed the trading parameters by entering “7” into the “basis points” data field. The trader/programmer then double-clicked the up arrow in an attempt to generate cancel/replace requests for all unexecuted orders with new limit prices that were 14 basis points higher than the original orders.
However, instead of replacing all unfilled orders with new orders with limit prices that were fourteen basis points higher, each click resulted in the generation of a separate set of cancel/replace requests. Below is an example of the double-clicking issue:
* Original limit order in XYZ with a limit price of P is submitted (Order A)
* Click #1 results in a replacement order in XYZ with a limit price of P + 7 basis points (Order A2)
* Click #2 results in another replacement order in XYZ with a limit price of P + 14 basis points (Order A3)
The change to the SmartWB’s order parameters resulted in cancel/replace messages for 38 of the 129 total securities being targeted by the SmartWB on November 14th. The majority of these cancel/replace requests were processed without incident and were ultimately filled if they were marketable or canceled if they were not. However, for the last seven of the 38 securities included in the cancel/replace requests, the cancel/replace request resulted in “reject-unmatched cancel” messages from SuperDOT, signifying that SuperDOT could not locate the orders that the SmartWB was attempting to cancel and replace. Significantly, the seven securities were all at the end of the “list” of the 38 securities for which the double-clicking resulted in cancel/replace requests.
Therefore, for every one of the final seven of the 38 securities impacted by the change to the algorithm’s order parameters, the following occurred::
* Original limit order in XYZ with a limit price of P is submitted (Order A)
* Click #1 results in a new limit order in XYZ with a limit price of P+7 (Order A2), which is intended to replace Order A
* Click #2 results in a new limit order with a limit price of P+14 in XYZ (Order A3) and attempts to replace Order A2 with Order A3 before SmartWB is able to send the request to replace Order A with Order A2.
* SuperDOT rejects the request to replace Order A2 with Order A3 because SuperDOT never received Order A2 and therefore cannot locate it.
Because SuperDOT was unable to locate the subject of these cancel/replace requests, SuperDOT responded by issuing “reject-unmatched cancel” messages. However, the SmartWB was not programmed to properly respond to the “reject-unmatched cancel” message and therefore continued to repeatedly re-send the cancel/replace requests.
Here comes the money:
By the end of trading on November 14th, the SmartWB had sent approximately 600,000 cancel/replace messages. SuperDOT sent approximately 405,000 “reject-unmatched cancel” messages between 3:40 p.m. and 4:05 p.m. in response to cancel/replace requests generated by the SmartWB. The seven affected securities traded at five different trading posts on the NYSE Floor. Message traffic in all of the securities traded at these five Floor posts either slowed considerably or stopped completely.
A whole stock exchange stopped.
All the other computers, which trade against idiot Joe Schmoe investors and each other, battling to trade that 0.00005 of a second faster... Dead. Stop.
For those of you with ADHD, the sequence of events, simplistically, was that in the mad trading rush prior to close (an event we have all noticed numerous times in Zero Hedge), the CS algo began trading per default parameters, which were subsequently changed by a trader, who however inadvertently generated a separate set of cancel and replace request, which in turn caused the NYSE to respond with 405,000 "reject-unmatched cancel" messages. This burst of activity clogged messaging traffic in five trading posts to stop completely, preventing trading to be completed by the normal closing time of 4 p.m., and in fact taking as long as 4.27 p.m. Amusingly, CS never knew anything wrong had happened and was not notified it snarled trading in numerous stocks until the next day.
But wait, there is more.
What about a government computer trader?
Is The SPY Getting A "Jump" At Key Levels From A Quant Algo?
The SPY is the index fund, "SPDR S&P 500 ETF (the Fund), formerly SPDR Trust, Series 1, generally corresponds to the price and yield performance of the S&P 500 Index." (This quote is not from the article, but from the fund official description. Basically it mirrors how the market is doing, think DOW JONES index).
Now for the juicy stuff...For months now i have been watching a specific algorithm push our markets around with great ease. It looks like this algo is giving the SPY a little push through support and resistance levels with massive size executed in seconds.
Resistance levels is a term that traders (of human variety) are obsessed about. Generally a stock or a commodity will trade between an upper and lower boundary, and most of its swings in price occur in between these boundaries - resistance levels.
Now, traders (of human and electronic kind) look for patterns - if a thingamajig stock is trading for the last 3 years between $3.00 to $4.25 (adjusted for inflation, say), they will buy when it is at near $3 (or below) and sell when it approaches $4.25.
Now think that the geniuses who programmed these algos trading supercomputers are mainly dealing with resistance levels input into the computer formula (Forget the sophisticated news scanning - this scenario of resistance level based trading is perhaps more plausible for much of computerized stock trading).
So what happens when thingamajig stock nose dives and goes down to $3.00, then to $2.90, then $2.89, then goes down and down and down....?
Humans and algos react the same way: SELL! It's a run! (Of course algos react a bit quicker, making Joe Schmoes sitting in front of their PC's crying onto their keyboard).
Panic, pandemonium, burning cars, revolution, etc etc.
Now, what this anonymous guest blogger on zerohedge is saying is that he (or she) spotted a pattern of a computer algo executing massive trades to prop up stocks.
(It has to be a computer program because, remember, algos act in fractions of a second, and that run on the stock could occur in about 60 second interval - or less).
Sometimes the push is tens of thousands of shares, the size all depends on the natural volume around the level which the SPY is trading at the time it may need a "Jump". For instance if the market is oversold on a 1 min time frame and is trying to break higher off lows but just cant get the party going on its own, the algo will come in and take offers until day traders, scalpers, swing traders jump in and chase the market higher.
Once the price gets "jumped" the algo just sits and waits till natural buyers and sellers are few and far between and it either dumps or takes in more. Usually the program will reset itself after a trade, then will wait till it senses low volume once again.
For some concrete evidence of this action i have done a quick illustration, which includes Time & Sales which only display prints on the exchange the algorithm does business on. This exchange is used because of its very nice rebate structure, and it allows the algo to exploit the SOES, meaning it cannot trade in blocks larger than 10000 shares per order. So what does it do, it takes blocks almost 10,000 shares multiple times a second, this price action causes the market to lift violently. This is not small money, remember small money follows big money.
The algo in question starts buying at 110.04 with one block of 9999 shares, followed by 60k more shares all bought in under two minutes. You can see from the chart how the SPY reacted, it violently moved higher all the way up to 110.55, where the algo dumped just about all of the shares, you can see the prints in the "dump" prints window, again only showing the print from the exchange the algo does business on. The algo did its job, the cash market snapped back, the components again caught a bid and moved higher through resistance. I.E. they look alive and well... Natural buyers came in above the 110.55 level chasing the market up another 50 cents or so before they left and the SPY fell again because the volume was not there to support the massive run up which took place over 15 minutes. As you can see the algo works in two capacities, it manipulates the market to the upside along with keeping S&P500 components trading in a liquid orderly "non flat" fashion.
So... how is the recovery going?
Well, the market is up, so... that means...
To all my (human) dinky investors trading on their rickety PC's through an online service - you are all idiots.