Skynet and the "Flash" Computer-Trading Monster!

== Why a Transaction Fee Might Save Us From The Terminator! ==

Here’s a vital issue under discussion (at last) on both sides of the Atlantic.  Governments, both rich and poor, urgently need two things: a way to calm speculation in the financial markets and also new ways to raise revenue. In late September 2011, the European Commission proposed a tax or fee on financial transactions. This appears to be part of the newly announced European Union plan, with Britain the sole dissenter.

"A levy of just 0.1 percent -- or even just 0.05 percent -- levied on each stock, bond, derivative or currency transaction would be aimed at financial institutions’ casino-style trading, which helped precipitate the economic crisis. Because these markets are so vast, the fee could raise hundreds of billions of dollars a year - from the sector of the economy that made towering profits while being directly responsible for our present depression, " writes Philippe Doust-Blazey in the New York Times.

Read the article.  But note that it does not mention the top reason for such a tax!  That it might benefit real human investors by slowing finance and equity trading back down to the speed of human thought.

Would that necessarily be a good thing? The concocted rationalization you will hear, in opposition to this proposal, is called “market efficiency.” According to what’s become a bona fide cult, any process or innovation that allows ever-smaller increments of trade to happen ever-faster is “efficiency,” and that will automatically lead to better allocation of society’s capital, and thus a skyrocketing economy.

This is wrong in many ways, starting with the pure fact that the flourishing of fast-cybernetic trading has directly correlated with the steepest decline in the health of capital markets in a century. Indeed, the increase in market volatility that we have seen lately, with sudden spikes in apparently random directions, can be generally attributed to this trend.**

== The Core Advantage of Having a Seat ==

Then there is the small matter of "seating." Historically, stock markets (and bourses for commodities and credit instruments) began with some fellows gathering at an inn or under a tree to regularly sell or trade share certificates to each other, or on behalf of other folks. When this became big business and moved into dedicated buildings, these Members held "seats" and practiced the standard methods of guilds everywhere, limiting who could come to the table and buy or sell. All outsiders had to hire seated members to trade for them, while paying commission.  Collusive club practices that have been banned (as recommended by Adam Smith) in most other parts of the economy!

Yes, a little market competition arrived recently, as E-Trade and other online member-brokers slashed commission prices for retail folks like you and me. New electronic methods made that possible. But also, the sheer volume of retail trading increased greatly, making up the difference.

What the seated members don't want us to notice is that flash-computerized trading gives them a new, incredible insider advantage over the rest of us. Even if you or I had a computer and program as good as the systems owned by Goldman-Sachs, we could never do what they do, because they are on the inside, making millions of flash trades for free!

You, on the other hand, would pay commission on every flash trade and - no matter how advanced or clever or wise your program - you would lose.  In other words the transaction tax already exists. It is levied by members of an elite cabal who use it to prevent anybody else from doing what they can do with new technology.

This is not "market efficiency." It is the kind of market warping influence manipulation that Adam Smith despised.

But absolute refutation of the "efficiency" argument comes from a different direction. From Physics, biology and thermodynamics.

== Computers and markets emulate life ==

Living creatures thrive by finding a steep gradient of usable energy. Green plants utilize the fact that incoming sunlight is thermodynamically clean and much less entropic than the surrounding environment.  (Greenhouse retention of Earth’s infrared radiation is thus intrinsically entropic.)

Some of this gradient is used by the plant to grow and reproduce, or else gets stored away, while some is lost as transaction cost.
Animals in turn consume plants for that stored useful energy, investing the time and effort to bite and chew and digest in order to benefit from some of the remaining gradient. Predators then pounce and bite and digest in the next stage. There are always losses with each transaction, hence the number of predators who can be supported at each scale gets smaller and smaller.

Along the way, each plant, herbivore and carnivore has parasites, intestinal worms and bacteria, etc., that grab some of that stored energy along the way. If they grab too much, the animal can’t get a steep enough or plentiful enough energy gradient and it dies.

Can you see it yet?  Beyond a certain level, increasing the total number of transactions does not make living systems more efficient.  It flattens all energy gradients and makes life unhealthy... even dead.

Oh, I can hear the objections! "Biology has no relevance to economics and finance."  Well, we could argue about that endlessly, but it doesn't matter. Because entropy - indeed, all of the things I described above, like energy gradients - are vital factors in modern information theory. And information theory is the exact basis that fast-traders claim to have underlying everything they do.

== Returning to capitalism ==

Yes, the existence of a stock market does create a habitat-ecology for living companies to compete, to form alliances, to prey on each other, to seek out the capital they need in order to grow.  Stock markets are vastly over-rated in the latter category, since the companies themselves benefit very little from wild swings in share price, except when offering NEW shares. (There should be substantial difference in the capital gains tax for new shares than for gains in the simple trading of old shares, an activity that only helps capitalism at very low efficiency.)

Still, you get synergies.  When a human investor looks at a company’s new product and bets “this new gizmo or service is gonna go big!” and orders 1000 shares on E-Trade, at low commission, then there’s a good chance that capital will flow to a place that can benefit and utilize it, all motivated by the hope of a nice profit “meal.”  Buyers always believe that the "true" value of the share is higher than the seller thinks, and yes, one of them will be proved wrong. That is the darwinistic aspect of equities markets. And under certain conditions - when all the players are getting fair access to all the information that they need and nobody's doing insider deals - it is good.

Defenders of computerized flash programs that dive in and pounce on any detected market trend, making millions of automatic trades, detecting or anticipating the decisions of human traders... these folks tout that they provide a service -  “efficiency.” But which efficiency?  For whom?

(Indeed, a whole new transatlantic fiber cable is being laid, just to enable a few brokerage firms to gain a couple milliseconds advantage; they’ll make billions. Do you see how that helps our market economy? I can’t.)

Here is the core article of faith among those pushing flash trading, as expressed to me in a note from a senior Wall Street partner . "There is a gap between the perceived value of the stock as desired by the buyer and perceived by the seller. That means the true and correct value isn't being accurately pegged. By jumping in between buyer and seller, we help eliminate this gap, causing the inefficient disparity to go away, helping the market settle on the true price."

Yee-god. Did you grasp that? He admits that he makes his living by leaping between buyer and seller, snapping up the value gradient that motivated the buyer to make an offer in the first place. And he manages to rationalize that he is doing a gooood thing. Good for buyer and seller. Good for markets. Good for capitalism. Good for himself.  Well... one  of the above.

Some have likened the ruthless voracity of these trading systems to the "tragedy of the commons," a very important concept you should be familiar with. Others suspiciously see the blizzard of computer-trading as a way to create a vast fog, behind which a few thousand oligarch golf-buddies can hide insider deals. Both views have some merit. But there is another that goes deeper to the heart-essence of the matter.

Want the exact parallel in nature? It is those gut parasites or e-coli or salmonella, or Typhus, who nibble away the gradient of potential profit that the human trader perceives, between the current asking price and what he or she feels the stock may soon be worth.  It is the core logic of parasitism. If you ever read Douglas Adams's The Hitchhiker's Guide to the Galaxy, these are the perfect passengers aboard the Golgafrincham B-Ark.

We need a small Transaction Fee not only because it will bring in revenue from the sector of the economy that made huge bonuses while wrecking our economy, restoring an emphasis on those providing new, competitively innovative goods and services.

The bigger reason is that human investors won’t care about - or even be aware of - a 0.1% trade fee.

But those computerized parasitical systems will howl in agony!  Thus, it will give you  a better chance to gain from your own savvy and insight, when you log into your E-Trade account.

Read: A Transaction Fee Might Save Capital Markets…and Protect Us From the Terminator!

== The Past and the Future ==

This kind of tax or fee is well-precedented.  Enacted to help pay for the first world war, the  "Securities Turnover Excise Tax" or STET worked just fine in the United States from 1914 to 1966. From 1960 to 1966, stocks were taxed at the rate of 0.1 percent at issuance and 0.04 percent on transfer. Bonds were taxed at the rate of 0.11 percent at issuance and 0.05 percent at transfer.

Indeed, there is a way that conservatives might find a way to rationalize a "fee" instead of a "tax," if all of the funds generated went first into paying the expenses of the SEC, the various insurance programs and other regulatory apparatus that they rely upon for the smooth functioning of their markets. (Thus removing these entities from the burdens shouldered by the public treasury.) Insurance funds could be greatly expanded to provide a cushion vs the next market contortion.  Especially, funds could be spent on balancing and fairness on the international scene.

In fact, there are some even in the Belly of the Beast who have come around to supporting a return to this sensible measure.  For example, David Harding, CEO of Winton Capital, $26B hedge fund, the world's largest quant-based fund, was  quoted in the Financial Times: "I would be in favor of a low [financial transaction tax] if part of it was used to finance more supra-national regulation of markets."

Moreover, it is already the norm in most advanced countries. The EU is apparently instituting a STET across the board in the Eurozone and Britain, whose "city" bankers rule the roost, nevertheless has a STET. For comparison, the UK's STET is  about .25 percent, and Taiwan just dropped theirs from .60 to .30 percent.

Nobel Prize winning economist Paul Krugman makes the case for reinstating the STET in more conventional terms than I do here:  "But wouldn’t such a tax hurt economic growth? As I said, the evidence suggests not — if anything, it suggests that to the extent that taxing financial transactions reduces the volume of wheeling and dealing, that would be a good thing.
 
"And it’s instructive, too, to note that some countries already have financial transactions taxes — and that among those who do are Hong Kong and Singapore. If some conservative starts claiming that such taxes are an unwarranted government intrusion, you might want to ask him why such taxes are imposed by the two countries that score highest on the Heritage Foundation’s Index of Economic Freedom."

Sen. Tom Harkin (D-Iowa) and Rep. Peter DeFazio (D-Ore.) recently introduced The Wall Street Trading and Speculators Tax Act that would impose a minuscule tax of 0.03 percent on financial transactions, meaning that longterm investors would barely notice it. Even so, it could raise more than $350 billion to reduce the budget deficit over the next nine years, according to an analysis by the Joint Tax Committee, a nonpartisan congressional scorekeeping panel.

== The Real Reason to Worry ==

So, what does all of this have to do with The Terminator? (Okay, I saved the sci fi perspective till last. But it has a terrifying kind of plausibility.)

You'll recall the by-now cliched premise of that film - it supposes that (sometime in the future) the U.S. military would develop a super computer/program/system called "Skynet" that gradually becomes self-aware through a process that theoreticians call "emergence from complexity." This is actually taken very seriously by deep thinkers about artificial Intelligence. Indeed, our first encounter with AI may come exactly that way... by surprise.

Only... supposing that this malevolent AI comes from a military source? That's pure Hollywood.  Such systems are built with high priority to systematic reporting, accountability, multiple redundancies, fail-safes and obedience to chain of command.  No, there are other complex computer systems that seem far more likely to suddenly become self-aware in powerfully dangerous ways.

Take those high-speed trading systems we've been discussing. They are growing incredibly sophisticated, at a very rapid rate, absorbing and incorporating models of human psychology, with one goal in mind. To appraise and predict behavior patterns in order for the program to track and to pounce on opportunities for predatory trading.  Competitive ferocity is the only criterion for success. Indeed, if you were to even propose inserting balancing factors like ethics or morality or accountability into such a project, you'd at-minimum be laughed down and probably fired.

A bizarre-sci-fi notion?  MIT researcher Alexander Wissner-Gross suggested that the first true AI could emerge on a planetary scale from the developing system of interlocked exchanges for high-frequency financial trading, which could be seen as a developing global “brain” already operating at relativistic speeds.

Moreover, these systems are receiving billions in funding (including their own new transatlantic fiber cable) entirely in secret.  There are no public agencies involved. No third party observers. No Congressional oversight committees.  No supervision whatsoever. Laboratories developing new genetic strains of wheat are under closer accountability than cryptic Wall Street think tanks that may unleash the first fully autonomous AI... programmed deliberately to have only the behavior patterns, goals, attitudes and morality of parasites.

And so we see the ultimate reason to demand the Transaction Fee. At a low level - say 0.1% - it would never bother a private citizen who is optimizing his portfolio on E-Trade, especially if each trader gets a hundred or so "freebies" that come exempt from the tax. But it would remove the  incentives that Wall Street "geniuses" now feel compelled by, to invest in these monstrous, hyper-fast trading programs that swamp the market in a blizzard of uncountable mosquito bites.

The fee (which could also help balance our budget) can be tuned to give that human a fighting chance and to discourage the very worst kind of artificial intelligence from leaping upon our necks out of the dark.

==Side Note: Gingrich and Asimov==

Both Republican former House Speaker Newt Gingrich and Nobel prize  winning Keynsian economist Paul Krugman have a trait in common.  They grew up fervent science fiction fans, especially transfixed by the future-historical speculations of Isaac Asimov.  Gingrich wrote about this influence that helped to shape his life.

“While Toynbee was impressing me with the history of civilizations, Isaac Asimov was shaping my view of the future in equally profound ways….For a high school student who loved history, Asimov’s most exhilarating invention was the ‘psychohistorian’ Hari Seldon.  The term does not refer to Freudian analysis but to a kind of probabilistic forecasting of the future of whole civilizations.  The premise was that, while you cannot predict individual behavior, you can develop a pretty accurate sense of mass behavior.  Pollsters and advertisers now make a good living off the same theory.”

Here’s an interesting essay comparing Gingrich’s obsession with Isaac’s Asimov’s compelling sci fi memes. Writes Ray Smock: "Edward Gibbon saw the decline of Rome, Hari Seldon saw the decline of the galactic empire, and Newt Gingrich saw the decline of America." Further, Newt was attracted to the idea of one man shaping the destiny of an entire civilization. Smock adds, "History and fiction seem more exciting when there is decline.  This gives heroes, visionaries, demagogues, and politicians something to fix."

As the similarly obsessed author of Foundation’s Triumph, who tied many of Isaac's loose ends, I'd have a thing or two to say to Newt! Could be an interesting intellectual tussle.

Still, this aspect to his background speaks well for him. Ten points.  Out of....

====== Notes on HFT ======

* In fairness, here is a report written by a finance industry think tank attempting to rebut the notion of transaction tax.  It appears to make precisely the "efficiency" arguments I predicted.  It appears, in any event, that the experiment will be run, if the new EU arrangements do include (as I've heard) a transaction fee.  One reason that the British government (largely swayed by the "City" banks and trading firms) opted out. Go ahead and hear the other side out.  Only remember, it all boils down to "supply side" mysticism that has never, ever seen any of its large scale predictions come true.    

**For example, when asked to appraise why markets experience a sudden crash on May 6, 2011, without any  apparent reason, Wall Street analyst Peter Cohan explained why such things are happening more often, and in wilder swings.

"... 70 percent of the volume of trading on the stock exchanges these days is done by something called flash traders, and that's basically computers that buy and sell stocks and hold them for about 11 seconds on average. So all of the discussions that we have the economy, politics, regulations, company earnings -- all that stuff -- there's just no way that a computer holding and selling a stock in 11 seconds is going to be able to do all of that analysis. So it's really all out the window. And there's really no clear-cut explanation for why stocks move up and down every day."

*** One expert wrote in, supporting the STET  with a series of sample portfolio experiments, showing that any human trader will see almost no inconvenience of "friction" but that flash predatory programs would lose their advantage: "Now compare this to the cost of the transaction taxes. In my own case, my most frequently traded portfolio is once a month. A .05% tax would amount to a 1.0005^12 or ~ .6% drag on my CAGR. For my quarterly and annual portfolios, the numbers are ~.2% and .05% respectively. I would point out that this is small compared to the reduction in drag that came from decimalization that happened after I started running this portfolio. 

"The useful thing is that this drag increases exponentially with trading frequency. Trade 1000 times and the drag is ~64%, 10,000 times and it is ~ 14,800%, a million times and my calculator overflowed. This means that even if you deem .6% too onerous for the monthly trader, you could go to .005% and the 64% bite occurs at the 10,000 trade level and 14,800% at 100,000 trades and now you can calculate that a million trades will give you a drag of ~5*10^23% all while costing me 0.06% a year on my monthly.
 

"You could even feed this money into the SIPC so that this provides additional counterpart protection as compensation for the minor sums being charged to the low frequency trader. I would argue that any putative efficiency gain from high-frequency fully automated trading is more than outweighed by the systemic risks of putting the financial system largely in the care of systems whose behavior cannot be comprehensively understood, much less predicted. That this can be limited while imposing essentially zero net cost on retail investors strongly argues for doing so."