"If only he'd used his powers for good-"

Baby brother is a True Crime fanatic.  I'm not sure how it entered our household, but the initiator, the spark that started the flame, had to have been the book "Bloodletters and Badmen" by Jay Robert Nash, an American encyclopedia of crime "from the Pilgrims to the present".  Baby brother had this book memorized before the age of probably ten. His appetite for the gruesome and bizarre has yet to be sated. I confess I also have a morbid curiosity towards the diabolically clever misdeeds, but my interest cannot hold a candle to his abiding obsession.

(I should note that baby bro also presented me with the Three Name Conjecture - that all successful assassins are referred to by their full name: John Wilkes Booth, James Earl Ray, Lee Harvey Oswald, Mark David Chapman, &c, &c. I pointed out that John Hinckley, would-be assassin of Ronald Reagan, is referred to as John Hinckley Jr., and his response was "How do you know Reagan wasn't assassinated and a lookalike replaced him"? Hmmm).

In any case, we somehow started talking about Dr. John Schneeberger. He managed to avoid arrest by spoofing a DNA test of his blood by implanting surgical tubing filled with another man's blood in his arm and fooling the forensic lab technician into sampling blood from the tube instead of his own veins.

Damn, that is a diabolical genius. To which baby bro offered up the TV trope "If only he had used his powers for good, instead of evil". To which I offered that I didn't think so. That this type of cleverness could only be used for evil purposes, and if tried to be cleverly benevolent, he would utterly fail. Or more to the ironic point of the Simpsons, he would be like Mr. Montgomery Burns, who is even more evil when tries to do good.

Which brings me to news which recently caught my attention, that the equity market is completely fucked up. And if I am looking for blame, I've got to point a finger at high frequency trading (HFT). Done in time span of milliseconds, these trades, carried out by sophisticated algorithms on supercomputers, are really nothing more than a form of parasitism upon financial networks. To those not familiar with the unsavory practice, it looks a lot like an artificial insertion of a middle man, a greasy Mob transaction, wherein "shipping and handling" is done for a markup fee without any true added value occurring.It is basically a form of extortion.

It goes like this. Most HFT transactions are made to "sniff out" the demand for a specific stock. So, ehre I am an institutional buyer, noting (in human time) that a particular stock looks a 'buy' or a 'buy and hold'. People start to buy, and the price of the stock goes up. The evil opportunistic algorithms operating on supercomputers, note this, issue buy orders for the stock and then, a millisecond later, cancel the transaction. They can do this, since 'their' microseconds are 'our' seconds, thousands of transactions per second. By ordering and cancelling a stock, they can determine what maximum market price much the buyer is willing to pay, and then they buy and sell the stock to the unwitting human. The brokerage using the HFT strategy thus pockets both the brokerage fee, and the articifical profit generate through the computer churn on the market.

It's an old scam. It goes back a long time, wherein a 'club' of buyers artificially bids up the price of something by incestuously selling it at increasing values between members of the club, until finally dumping the inflated stock onto the suckers, e.g the market. Almost immediately the price drops to nothing, leaving the consumer to absorb the costs of the entire transactional clusterfuck.

Interestingly, fully 60 to 70% of all transactions done on the US market are HFT transactions. Of those transactions, more than 95% are immediately cancelled. In total, only 2% of the 20,000 brokerages account for all HFTs, according to this article. Those are the vampire squid motherfuckers sucking on the face of humanity.

The consensus suspicion of market manipulation is that HFTs were responsible for the May 6th, 2010 Flash Crash, when the market dropped some 600 points, and then rebounded  some minutes later. Market volatility is now at an all time high, as seen in this article from Zero Hedge.

The worry now is that HFTs pose continual future risks of more crashes and spikes to come. So much so, that it garnered the attention of the physics community, as seen in this article: "Study links ultrafast machine trading with risk of crash".

The authors examined the price logs of over 60 markets from 2006 through 2011. They discovered what they termed "fracture events" within the market during this period, some 18, 520 ultrafast black swan events". Given the speed at which they occur, it is impossible for human supervision to control these events, and black swans have a nasty tendency to turn into Dragon Kings.

Since the Flash Crash, the SEC has implemented circuit breakers, but many fear this is not enough. One suggestion is for the implementation of filters in traffic that would slow down opening transactions but not closing transactions, which would hinder a price collapse. Another suggestion is to keep orders exposed for a full second, to accommodate the glacially slow human reaction time. This would reduce the volume of HFT transactions, but not halt them. Still another suggestion is introduce (or rather, reintroduce) a small "Tobin tax" upon all share transactions, thus making most HFT scams unprofitable. The latter suggestion seems eminently doable, and makes the most sense.

Nevertheless, I took it upon myself to read the paper behind this article ("Financial black swans driven by ultrafast machine ecology", http://arxiv.org/abs/1202.1448, if interested), and was struck with the fact that there are algorithmic ways of preventing fractures:

"Just prior to a large change, the resulting system is therefore (momentarily) largely deterministic. If left alone, it will produce a large change in a definite direction - however, it can be driven away from the impending crash or spike by increasing the strategy diversity in the following ways, listed here in order of increasing effect but also increasing assumed knowledge: (1) if little is known about the system, our earlier expressions for η < 1 predict that increasing the disorder in the initial strategy allocations by adding agents with randomly chosen strategies will reduce the scale of fluctuations toward their lower bound..."

 Okay, okay, there are two other methods listed, but the operative phrase here is insert new agents. In other words, have an computerized algorithmic army out in the transaction space, spiders, searching for unusual transactions, who, when they detect this type of scam, fuck with the HFT brokerage firm's computers.

How to do this? Well, conficker comes to mind. Ten million or so CPUs, strategically scattered throughout the globe, all of them fairly near optimal intermediate trading node locations, all running continuously in the background on their pirated PCs, ready to fuck up Wall Street's scams.

Antibodies, if you will, to be used against some particularly nasty viruses.

If not conficker, then, perhaps a botnet for the rest of us.

Now wouldn't that be cool?