Nobody knows if HFT is good or bad

The publication of Michael Lewis' new book Flash Boys: A Wall Street Revolt has pushed discussion of high frequency trading (HFT) to the fore.

Noah Smith opines that nobody knows if HFT is really good or bad.

Do market-makers increase or decrease liquidity? Do front-runners increase or decrease it? What about informational efficiency of prices? What about volatility and other forms of risk, at various time scales? What about total trading costs? Good luck answering any of these questions. Actually, Stony Brook people are working on some of these, as are researchers at a number of other universities, but they are huge questions, and our data sets are incredibly limited (data is expensive, and a lot of stuff, like identities of traders, just isn't recorded). And keep in mind, even if we did know how each of these strategies affected various market outcomes, that wouldn't necessarily tell us how the whole ecosystem of those strategies affects markets - after all, they interact with each other, and these interactions may change as the strategies themselves evolve, or as the number and wealth of the people using each strategy changes. 

Confused yet? OK, it gets worse. Because even if we did know how HFT affects markets, we don't really know if it's good or bad on balance. For example, HFT defenders often say HFT provides "liquidity". Is liquidity good for markets? How much is liquidity worth, are there different kinds of liquidity, and does it matter when the liquidity comes? If I have a bunch of totally random trading, that certainly makes markets liquid, but is that a good thing? Actually, maybe yes! In lots of models of markets, you need random, money-losing "liquidity traders" in order to overcome the adverse selection problem, thus inducing informed traders to trade, and getting them to reveal their information. But HFTs don't lose money, they make money - is their liquidity provision worth the cost?

To know that, even if we knew the impact of HFTs on informational quality of prices, we'd have to know the economic value of informational efficiency. Suppose the true worth of GE stock. according to the best information humanity has available, is $100. Suppose the price is $100.20. How bad is that? How much is it worth, in economic terms, to push the price from $100.20 to $100.00? Is it worth $0.20 per share? It depends on how GE's stock price affects the company's investment decisions. To know that, we need an economic model of corporate decision-making. We have many of these, but we don't have one over-arching one that we know works in all circumstances. Corporations are way more complicated than what you read in your intro corporate finance textbook!

(And this is all without thinking about weird things like behavioral effects of the humans who interact with HFTs...)
 

I don't know much about HFT other than a few articles I've read here or there, so count me among those who have no idea if it's positive or negative.

Procrastination

A really wonderful two-part series on procrastination over at Wait But Why:

  1. Why Procrastinators Procrastinate
  2. How to Beat Procrastination

I hate patents, but the term Dark Playground perhaps deserves a trademark.

The Dark Playground is a place every procrastinator knows well. It’s a place where leisure activities happen at times when leisure activities are not supposed to be happening. The fun you have in the Dark Playground isn’t actually fun because it’s completely unearned and the air is filled with guilt, anxiety, self-hatred, and dread. Sometimes the Rational Decision-Maker puts his foot down and refuses to let you waste time doing normal leisure things, and since the Instant Gratification Monkey sure as hell isn’t gonna let you work, you find yourself in a bizarre purgatory of weird activities where everyone loses.

Esther Perel on infidelity

In America, infidelity is described in terms of perpetrators and victims, damages and cost. We are far more tolerant of divorce with all the dissolutions of the family structure than of transgression. Although our society has become more sexually open in many ways, when it comes to monogamy, even the most liberal minds can remain intransigent. When discussing infidelity, we use the language of moral condemnation. And it isn’t only the act that’s reprehensible; the actor, too, is judged by the strictest standards. Adultery becomes a moral failing as we move to a description of character flaws: liar, cheater, philanderer, womanizer, slut. In this view, understanding an act of infidelity as a simple transgression or meaningless fling, or a quest for aliveness is an impossibility.

An affair sometimes captures an existential conflict within us: We seek safety and predictability, qualities that propel us toward committed relationships, but we also thrive on novelty and diversity. Modern romance promises, among other things, that it’s possible to meet these two opposing sets of needs in one place. If the relationship is successful, in theory, there is no need to look for anything elsewhere. Therefore, if one strays, there must be something missing. I’m not convinced.

...

The current view is that infidelity depletes intimacy and is a breach of trust and commitment, both emotional and sexual, that can never be fully recouped. Even the psychological literature focuses almost exclusively on the ravages of infidelity. I’d like to offer a view that challenges this premise and encompasses both growth and betrayal at the nexus of affairs.

Though affairs often result in deep emotional crisis, deception and betrayal are not the prime motivation. I suggest we look at infidelity in terms of growth, autonomy, and the desire to reconnect with lost parts of ourselves. Perhaps affairs are also an expression of yearning and loss.
 

Lots more here from Esther Perel on infidelity, all of it fascinating. Though I've never been married or had an affair, this passage had a ring of familiarity:

Sometimes, we seek the gaze of another not because we reject our partner, but because we are tired of ourselves. It isn’t our partner we aim to leave, rather the person we’ve become. Even more than the quest for a new lover we want a new self.
 

The pressure on the institution of marriage is higher than it's ever been because we now expect our spouses to provide so many different forms of fulfillment. In an interview with Slate, Perel notes:

What’s changed is, we expect a lot more from our relationships. We expect to be happy. We brought happiness down from the afterlife, first to be an option and then a mandate. So we don’t divorce—or have affairs—because we are unhappy but because we could be happier. 
 

This reminded me of two of my recent posts about the new definition of marriage: marriage is now all-or-nothing, and hedonic marriage.

Modern movie studio economics

A really fantastic three part analysis by Liam Boluk of modern movie studio economics as it pertains to blockbusters.

Future of Film I: Why Summer 2013 was Destined for Losses

Much has been said about the growing role of ‘tent-pole’ filmmaking, where the superlative performance of a major blockbuster supports the rest of the studio’s portfolio (including failed blockbusters). In practice, however, the strategy doesn’t ‘hold up’. Over the past decade, the Summer Blockbuster season has delivered a net theatrical profit only three times and the major studios have lost nearly $2.6B on $34B in production and marketing spend.

...

The Summer 2013 season was so jam packed with “blockbusters” that the industry seemed destined for historic losses containing:

  • 18 blockbusters – A historic high and 41% increase over the ten year average
  • 15 back-to-back weekends of blockbuster releases – A third more than ten year average and 25% more than a decade ago
  • 5 weekends with two blockbuster releases – 317% more than the average, 2.5x the previous record and 5x the number in 2003
     

Future of Film II: Box Office Losses as the Price of Admission

For all its glamour, theatrical entertainment is simply a rotten business to be in.

Though its products are not commodities, many of the industry’s competitive dynamics and characteristics suggest they could be:

  • Past success is not a predicator of future performance. Last year’s box office receipts do not influence current-year performance and year-to-year momentum translates into little beyond high spirits
  • Talent doesn’t ensure success. The most “valuable” stars, brand-name directors and veteran producers routinely produce box-office bombs
  • Hollywood brands are irrelevant. Aside from Pixar (whose brand is arguably in decline), consumers don’t pick films based on whether they were a Universal or Paramount production. Indeed, consumers rarely even know
  • All products are offered at the same market price. Regardless of the film’s production costs or target customers, end consumer pricing is largely identical

...

Why then, do executives continue making films? They have few (if any) levers they can reliably play with, the success of individual films causes massive disruptions in annual performance and in the long run, performance is unlikely to break-even, let alone outpace market returns.

The answer: ancillary revenue. In 2012, box office receipts represented only 52% of revenue for the average film, with the remainder comprised of home video sales, pay-per-view and TV/OTT licensing, syndication fees and merchandising. After appropriating for related costs, as well as backend participation (Robert Downey Jr. took a reported $50M from Avengers) and corporate overhead, the average Internal Rate of Returns (IRR) for the majors jumps to roughly 80%.

...

Since silent films first appeared on the silver screen, motion pictures has been primarily a B2C business, with film studios sharing revenue with theater operators. But over the past decade, the majors have transformed into an increasingly diversified B2B partner. Their job is not to bring eyes to their theatrical products, but to enable NBC to drive Sunday advertising revenue, ABC Studios to create a high-margin television series, HBO to collect monthly subscriber fees or Mattel to sell Cars toys. Entertainment, in short, has become both a platform and a service.
 

Future of Film III: The Crash of 'Film as a Platform'

More important, however, is the impending ‘Film as a Platform’ implosion. Looking at 2016′s dense release schedule, theatrical losses per blockbuster are likely to increase considerably. Not only will increased competition drive down average attendance, it could push studios to invest even more into their film properties in the hopes of standing out. This itself isn’t a fatal exposure – studios will simply need to rely more heavily on ancillary revenues. However, the real issue is that further audience fragmentation will make it even harder to achieve the critical mass audience needed to support ancillary revenue streams. Worse still, the growing number of franchise films may end up flooding ancillary channels.

Ancillary markets such as home video, merchandising and children’s television can only absorb so much content. A child, after all, will not want a Christmas comprised of various X-Men, Star Wars and Avatar paraphernalia and parents are unlikely to purchase multiple bedroom sets. Television audiences, can support only so many series in a given genre (the Marvel Cinematic Universe will have 7 in 2015 alone). Though themed sets have been a strong sales driver for the Lego Group, optimizing marketing and inventory investments will limit the number of franchises they will support – especially in the holiday season. As a result, the deluge of ‘platform films’ is likely to significantly reduce the ancillary revenues studios rely on for film profitability. To make matters worse, it would take at least two years for studios to emerge from this crunch due to the fact films are released 1-2 years after investment/production decisions are made.
 

I love my occasional summer blockbuster movie, but I already feel like I have pop movie diabetes. The latest Captain America movie has not one, but two extra scenes during the end credits, each previewing a different future Marvel movie. One day soon, after the credits of the latest Marvel movie, the extra scene will just be a house ad for the theme park ride based on the movie, and on the way out of the theater there will be a booth set up to sell toys from the movie. Theaters already make most of their profits on concessions, using the blockbuster movies as a loss leader, it's not all that different for the studios themselves.

Notch one more for computers

In May, 1997, I.B.M.’s Deep Blue supercomputer prevailed over Garry Kasparov in a series of six chess games, becoming the first computer to defeat a world-champion chess player. Two months later, the Times offered machines another challenge on behalf of a wounded humanity: the two-thousand-year-old Chinese board game wei qi, known in the West as Go. The article said that computers had little chance of success: “It may be a hundred years before a computer beats humans at Go—maybe even longer.”

Last March, sixteen years later, a computer program named Crazy Stone defeated Yoshio Ishida, a professional Go player and a five-time Japanese champion. The match took place during the first annual Densei-sen, or “electronic holy war,” tournament, in Tokyo, where the best Go programs in the world play against one of the best humans. Ishida, who earned the nickname “the Computer” in the nineteen-seventies because of his exact and calculated playing style, described Crazy Stone as “genius.”
 

Computers overtake humans in yet another field. After Deep Blue prevailed over Kasparov in chess, humans turned to the game of Go for solace. Here was a game, it was said, that humans would dominate in for quite some time. It turned out to not be much time at all.

Coulom’s Crazy Stone program was the first to successfully use what are known as “Monte Carlo” algorithms, initially developed seventy years ago as part of the Manhattan Project. Monte Carlo, like its casino namesake, the roulette wheel, depends on randomness to simulate possible worlds: when considering a move in Go, it starts with that move and then plays through hundreds of millions of random games that might follow. The program then selects the move that’s most likely to lead to one of its simulated victories. Google’s Norvig explained to me why the Monte Carlo algorithms were such an important innovation: “We can’t cut off a Go game after twenty moves and say who is winning with any certainty. So we use Monte Carlo and play the game all the way to the end. Then we know for sure who won. We repeat that process millions of times, and each time the choices we make along the way are better because they are informed by the successes or failures of the previous times.”

Crazy Stone won the first tournament it entered. Monte Carlo has since become the de facto algorithm for the best computer Go programs, quickly outpacing earlier, proverb-based software. The better the programs got, the less they resembled how humans play: during the game with Ishida, for example, Crazy Stone played through, from beginning to end, approximately three hundred and sixty million randomized games. At this pace, it takes Crazy Stone just a few days to play more Go games than humans collectively ever have.
 

Well, at least we still have Arimaa.

The trolley problem and self-driving cars

The trolley problem is a famous thought experiment in philosophy.

You are walking near a trolley-car track when you notice five people tied to it in a row. The next instant, you see a trolley hurtling toward them, out of control. A signal lever is within your reach; if you pull it, you can divert the runaway trolley down a side track, saving the five — but killing another person, who is tied to that spur. What do you do? Most people say they would pull the lever: Better that one person should die instead of five.
 
Now, a different scenario. You are on a footbridge overlooking the track, where five people are tied down and the trolley is rushing toward them. There is no spur this time, but near you on the bridge is a chubby man. If you heave him over the side, he will fall on the track and his bulk will stop the trolley. He will die in the process. What do you do? (We presume your own body is too svelte to stop the trolley, should you be considering noble self-sacrifice.)

In numerical terms, the two situations are identical. A strict utilitarian, concerned only with the greatest happiness of the greatest number, would see no difference: In each case, one person dies to save five. Yet people seem to feel differently about the “Fat Man” case. The thought of seizing a random bystander, ignoring his screams, wrestling him to the railing and tumbling him over is too much. Surveys suggest that up to 90 percent of us would throw the lever in “Spur,” while a similar percentage think the Fat Man should not be thrown off the bridge. Yet, if asked, people find it hard to give logical reasons for this choice. Assaulting the Fat Man just feels wrong; our instincts cry out against it.

Nothing intrigues philosophers more than a phenomenon that seems simultaneously self-evident and inexplicable. Thus, ever since the moral philosopher Philippa Foot set out Spur as a thought experiment in 1967, a whole enterprise of “trolley­ology” has unfolded, with trolleyologists generating ever more fiendish variants.
 

There are entire books devoted entirely to the subject, including the humorously titled The Trolley Problem, or Would You Throw the Fat Guy Off the Bridge: A Philosophical Conundrum or the similarly named Would You Kill the Fat Man?: The Trolley Problem and What Your Answer Tells Us about Right and Wrong. If the obese don't have enough problems, they also stumble into philosophical quandaries merely by walking across bridges at inopportune moments.

In the abstract, the trolley problem can seem frivolous. In the real world, however, such dilemmas can prove very real and complex. Just around the corner lurks a technological breakthrough which will force us to confront the trolley problem once again: the self-driving car.

Say you're sitting by yourself in your self-driving car, just playing aimlessly on your phone while your car handles the driving duties, when suddenly a mother and child step out in front of the car from between two parked cars on the side of the road. The self-driving car doesn't have enough time to brake, and if it swerves to avoid the mother and child, the car will fly off a bridge and throw you to certain death. What should the car's driving software be programmed to do in that situation?

That problem is the subject of an article in Aeon on automated ethics.

A similar computer program to the one driving our first tram would have no problem resolving this. Indeed, it would see no distinction between the cases. Where there are no alternatives, one life should be sacrificed to save five; two lives to save three; and so on. The fat man should always die – a form of ethical reasoning called consequentialism, meaning conduct should be judged in terms of its consequences.

When presented with Thomson’s trolley problem, however, many people feel that it would be wrong to push the fat man to his death. Premeditated murder is inherently wrong, they argue, no matter what its results – a form of ethical reasoning called deontology, meaning conduct should be judged by the nature of an action rather than by its consequences.

The friction between deontology and consequentialism is at the heart of every version of the trolley problem. Yet perhaps the problem’s most unsettling implication is not the existence of this friction, but the fact that – depending on how the story is told – people tend to hold wildly different opinions about what is right and wrong.

Pushing someone to their death with your bare hands is deeply problematic psychologically, even if you accept that it’s theoretically no better or worse than killing them from 10 miles away. Meanwhile, allowing someone at a distance – a starving child in another country for example – to die through one’s inaction seems barely to register a qualm. As philosophers such as Peter Singer have persuasively argued, it’s hard to see why we should accept this.
 

If a robot programmed with Asimov's Three Laws of Robotics were confronted with the trolley problem, what would the robot do? There are long threads dedicated to just this question.

Lots of people have already foreseen this core ethical problem with self-driving cars. I haven't seen any consensus on a solution, though. Not an easy problem, but one that we now have to wrestle with as a society.

Or, at least, some people will have to wrestle with the problem. Frankly, I'm happy today when my Roomba doesn't get itself stuck during one of its cleaning sessions.