Efficiency: Sometimes More is Less

Listeners of the Radiolab podcast might recall an episode earlier this year about how the stock market works.  For those who do not listen, Vice’s Motherboard blog  (URL below) provides a handy catch-up.

The advent of High Frequency Trading (HFT) has the potential for extreme market disruption, as the world witnessed on May 6, 2010.  A mutual fund firm out of Kansas placed a large sell order for a particular security, and HFTs–which can conduct billions of transactions in a single day–precipitated a “Flash Crash” wherein the Dow dropped by almost 1000 points, the largest intraday loss in history.The HFT algorithms, which were unable to distinguish between an impending crash and a large, one-off order, began automatically selling and buying back securities at volumes and prices that  had no real basis in the market.


Some European countries have responded to the advent of High Frequency Trading by introducing the one thing that is sure to cause “friction” in an otherwise “efficient” process–a tax.  Both the French and Italian governments have passed laws that tax acquisitions and dispositions of covered securities in an attempt to reduce the effect of high frequency trading, but it is questionable whether such a thing would ever happen in the United States.  The idea is basically that instead of opting to conduct millions of trades to exploit price differences of as little as $0.0001, traders will balk at the additional tax liability and record-keeping burden that attaches to each exchange.  Even Americans, who may trade in American Depository Receipts (essentially securities that represent shares traded in French or Italian markets) will pay taxes to these foreign governments in the form of fees that are passed on to them from the in-country subcustodians.


Given Americans’ typical aversion to new taxes, what are our options?  The SEC has partially responded by introducing rules against fraudulent trading, which can provide an after-the-fact enforcement of rules designed to prohibit such practices, but the additional time and expense of conducting investigations and prosecutions renders it a slower and less effective (?) method of deterring high frequency trading.


What do you think are the best ways to approach the problem?  SEC rule-making? Levying a transaction tax?  Or is it even a problem, given that some proponents believe that high frequency trading provides liquidity to the market?



Mother Jones Article (includes Radiolab podcast):



SEC Findings Regarding the Flash Crash:



Information on the Italian Transaction Tax:


High Frequency Trading:



SEC: Skynet…E-something…Cyberdyne?

An early appearance of a Terminator machine in American court history is the familiar case of Kaltko v. Briney.  As you may recall from first year Con Law, this is the case where an enterprising couple from Iowa set up a trap to deter maim or kill trespassing thieves.  When a would-be burglar triggered the trap, which was deliberately aimed “to hit the legs,” most of his leg was blown off.  He sued for damages and eventually won at the Iowa Supreme Court.

Twenty-seven years later, the first movie in the Terminator series was released.  The important similarity is not how John Connor also instructed his shotgun-wielding death machine to aim for the legs, but in a larger plot point:  in the series, a worldwide nuclear apocalypse (“Judgement Day”) is initiated when an automated defense system, Skynet, interprets a command to take it offline for maintenance as an attack, and releases the American nuclear arsenal on Soviet Russia.  The Soviets respond in kind and pretty much everyone gets vaporized.  Obviously, there was some ret-conning in order for later sequels as the USSR no longer existed in 1997, the year of the apocalypse.

Some actors in the securities markets apparently have not heeded either the legal precedent of Kaltko nor the fictional cautionary tale of Terminator and have invested several billion dollars in high frequency trading supercomputers that buy and sale securities at a staggering pace—billions of transactions per second. 

One result of this trend apparently has been the “Flash Crash” of 2010, wherein the Dow plunged by almost 1,000 points within a single day.  Some investigations are still ongoing, but the SEC reporting so far has indicated that a single, massive sale of E-Mini contracts by a Kansas-based mutual fund sent the High Frequency Trading Algorithms into a tizzy, selling off enormous amounts of shares and driving their prices down rapidly.  By the end of the day, buyers had snapped up the egregiously undervalued shares and the market closed just 3% down.  A big sale by a mutual fund company isn’t quite the same as a nuclear attack, but the HFTAs—like Skynet—hadn’t been programmed to distinguish between a large transaction and an impending liquidity crisis.

If you were trying to liquidate part of your portfolio to, say, make a down-payment on your house or pay for your kid’s tuition bill, and your order happened to be executed during certain hours of the day, you might have been pretty disappointed—like the man whose investment wasn’t so much liquidated as vaporized when he lost $17,000 selling a blue chip stock. 

The SEC’s stated mission is “to protect investors, maintain fair, orderly, and efficient markets, and facilitate capital formation.”  The modern notion of market efficiency, an economic model that relies heavily on frictionless transactions, ironically emerged in the 1960s while at the same time capital markets were losing billions of dollars in what is known as the “Paper Crunch.”  Securities transactions were so inefficient (all being done via delivery of physical certificates) that the backlog trades to linger before settlement for days and days at a time.  Versions of the Efficient Market Hypothesis have been enshrined in the fraud-on-the-market doctrine which has been successfully invoked at the Supreme Court as recently as the summer of 2012 in the Erica P. John Fund v. Halliburton case.

But efficient market theories often depend on an asset pricing model, such as the Capital Asset Pricing Model (CAPM) or others, that assume a risk-averse, buy-and-hold investment outlook.  Rates of return, or “future cash flows,” feature heavily in the models, but in the market today most of the trades are executed by computer programs that don’t plan on collecting any dividends.


On the Flash Crash of 2010:




On Asset Pricing Models:



SEC explains what it does: 



Probable Paws

As an enthusiastic dog-owner (for two weeks now), I enjoyed reading “Inside of A Dog: What Dogs See, Smell, and Know ” by cognitive scientist Alexandra Horowitz–I’m a sucker for airport bookshops.  The author describes how humans have six million olfactory receptor cells while dogs have not only many more–as many as three hundred million–but they also have a much more robust neurological apparatus for detecting and interpreting the nerve signals that they send on to the brain.

 So dogs smell better than we do, which is not news.  Humans, though, have a much better grasp of the Fourth Amendment’s protection against unreasonable search and seizure.  In 2013, the Supreme Court made some important decisions about the use of trained police dogs in establishing probable cause.

 Comparing the use of sniffer dogs to the unwarranted use of thermal imaging, Justice Scalia wrote an opinion in Florida v. Jardines holding that police officers may not bring a drug-sniffing dog to your doorstep and use a canine response to establish probable cause for obtaining a warrant to search the home.

 Article about 2013 Supreme Court Opinions Regarding Police Dogs: http://www.reuters.com/article/2013/03/26/us-usa-court-dog-sniffs-idUSBRE92P0NE20130326

Insufferably Enthusiastic Sharing of Pictures and Videos of My New Dog:Upon request.  (During the first week, a request was unnecessary).