Arbitraging the human psyche

Because, I suppose, of its umbilical connection to the requirements of everyday life, the financial world has engaged some deep psychological truths such as, for example, the "self-fulfilling prophecy" describing disastrous bank runs.

Now, high-speed trading has resolved the old joke about the poor financier walking ten miles to school each day and ten miles back--both uphill.

Of course most of us realize that high-frequency (or high-speed) trading in and of itself is not dangerous or even disingenuous. After all, the world succeeds in generating more wealth for all when market makers as well as scientists and educators and writers innovate and succeed. Traders using the "mysterious algorithms" and cutting-edge equipment needed for high-frequency trading claim with a vaporous wave that they "add liquidity" or some other benefit to the market. Naturally, though, we outsiders assume that we are being unfairly gamed since history confirms this again and again. Gamed--as in zero-sum gamed with us retail investors and pension funds as the opponent, unfairly--as in they play by rules both unarticulated and unavailable to mortal retail clients.

Investigators have looked at the obvious such as front-running and even the meta trade practice of channel clotting. (Front-running is when you unfairly get to look at the panel of potential trading partners and then cherry-pick the partner offering you the best price regardless of the time submitted of the order compared to your benighted customers; Channel clotting--a variety of denial-of-service attack--is to issue so many bids, asks, and cancellations that your exchange co-located conspirators get virtually the entire order channel to itself for crucial milliseconds in order to manipulate the market by removing unfavored liquidity--even if only for a few hundred milliseconds at a time.)

Yes, evidence of both those have been found. As well as another dozen of the old cheating saws that would have been extinct decades ago if the various securities overseers have had more incentive than even a congressman going to work as a lobbyist for the industry in which he or she oversaw. But that's another story.

But where again, I suppose, only a writer might look, the "mysterious algorithms" describing high-speed trading perhaps are not as mysterious as its first blush. The act of trading, even when accounting says the units out are equal to the units in, can arbitrage the human spirit to the financial advantage of the traders.

Let's take a simple example. XYZ corporation has been muddling along at $10 a share without much of a catalyst to make it move upward or downward. A PR statement from XYZ is issued. Before it's content is analyzed and disregarded as not particularly material to the value of the corporation--just some exercise in image-building--the high speed trader--call him Frankentrader--issues a hundred thousand sell orders of a few shares apiece. In this fictional example the total offered is say one percent of the daily market float of the stock. Within milliseconds new sell orders are issued by interspersed with cancellation orders for elements of the first block. Within a second or two the entire sets of trades has been mostly cancelled. Frankentrader's entire risk capital has gone from a million dollars to a few thousand or tens of thousands in this flurry of orders.

But to the human eye and mind, the stock took a momentary dip of half-a-point. With the hundreds of thousands of years of safety found in following the herd, we slow humans respond and perhaps sell a bit of stock ourselves from fear; further, other automated trading platforms amplify the first hundred thousand orders with their block of sell orders on a hysteresis delay that is also noticed by the human psyche only at human speeds. The stock dips as much as three-quarters of a point.

Frankentrader now issues buys of a few shares, gradually closing his position, interspersed within the interacting rings in the pond that he himself caused just seconds before. The stock price slowly ticks higher as they make purchases and close positions, but never as high as when they sold. At the end of less than a minute or so, Frankentrader has completely liquidated his position in XYZ and made thousands or tens of thousands of dollars, merely by exploiting both human evolution coupled by our organic speed limits. And there are 23.4 million milliseconds in a normal trading day. Fertile ground to stake high-frequency claims in this wild, wild west of nano-scale psychology.

But the secret is not the simple fact that computers are faster that people: you see, Frankentrader and the master analyst underneath the "mysterious algorithm" simply knows that an instantaneous drop of 10 basis points with a single human perceptual frame causes more fear in humans than ten serial gains of one basis point over several human perception frames causes corresponding elation. The math is commutative but the human reaction is not.

And that asymmetry is the true underlying basis of the perpetual wealth generating machine of high-speed trading.

A new math PhD applying for work with a high-speed trader would be best served by studying Dostoyevsky's The Gambler or perhaps even Crime and Punishment and an encyclopedia of optical illusions or sleight of hand, and thereby the consequence of encompassing the post-graduate axiom that money can indeed flow downhill to them both ways.

###

If you like the cut of my neural jib, buy my new technothriller on the coordinated Chinese cyberattack on the United States, and MITs shadow laboratory's response. Goods are checked at the border for weapons of mass destruction. Computer firmware--never.
http://amzn.to/R4ypa8

Straddling technological change

When traveling to Sierra Leone, I imagine the activities I saw such as growing food, constructing mud brick houses and tending donkey carts (without the donkeys because of the Tsetse fly) were pretty much the same as a hundred years ago. While likely fewer opportunities exist for Sierra Leoneans to adopt rapid technological change, the same has not been true in the West, especially for those embracing (or infatuated by) technology.

Training to be a software engineer in the late 1970's at The Evergreen State College, I recall designing and initial coding using yellow pads of paper and pencil. I would then transcribe to the TTY300 or, if I were luck, the Tektronix or ADM video terminal equivalent front ends for our wonderful HP 2000C time-sharing computer. The process worked out fine; I could code and debug outside in view of the clock tower.  A louche hippy. The needed productivity was low enough: "four lines of debugged code per day," that the paper tech was not the chokepoint. Besides, I was helping to write Civilization, which as the time in the early to mid seventies was the largest multiplayer role-playing computer game in existence. Productivity was not a problem; I would just work eighteen hours a day on it.

Then I worked for Apple. Productivity by my peers became more on the order of 50-60 debugged lines of code per day or higher. Because each engineer could now have their own Apple ][, there was no longer a need to use paper. As long as one could get over the transition from paper design to video design and coding, productivity took a sharp turn upward. We still had to hand carry our five-inch floppies for golden mastering by the nascent code-control library person while the AppleTalk network and server design was worked out at Bandley III and the rest of the Cupertino campus.

Now productivity for software coders is measured on the order of 200-300 debugged lines of code per day. However the entire stream is now and has been for many years, paperless. From initial notes, to design review slides, to management and coding, building and package art, through final production and customer fulfillment and backup, the entire process is electronic. Even my friend in pubs had to be happy with the end result an Acrobat file or a web page or two rather than a manual.

Creative writing has lagged. Even now I use a yellow legal pad to jot out initial ideas, timelines and plot outlines. It sits by one of my monitors as I type the elaboration of its storyline into a suitable back-lit keyboard on my desk. Playing with my fetish Pelikan fountain pen or Uniball vision elite was as important as a Gauloise was to Sartre arguing over the evocation of Nausea. I guess the cultural inertia of the belletrist is pretty high.

But having changed my reading from mostly paper to mostly e-reader years ago (from the early days of the greenish screen Palm Pilot) I am making the transition to the total electronic work-flow: new Apple iPad, Notability app, desktop or laptop, Kindle store, electronic downloads and payment, to ACH deposit of royalties and finally buying groceries with a debit card. I don't know how it will work out, but if history is the judge, probably pretty well, after the initial dislocation.

I am guessing the rule-of-thumb of a thousand words a day will fall to the new rule of three thousand words a day for us transducing axon depolarization to glyphs.

I wonder what will be next to improve creative productivity?

What I want is a direct author to reader link, as well as a continuous flow as the successive refinement of work in the process of creation--all the while actually making a living at it, if the quality of the craft and the ongoing contentment of the reader warrants it.

But before you ask, I am still keeping the Pelikan safe in its worn leather case as it always will be one of my life's friends.