Buying stocks by the speed of light, in the middle of the ocean
April 7, 2011 7:24 PM   Subscribe

"..the time it takes light to propagate between [stock] exchanges, for example between New York and London, is now a limiting factor in [financial] trading.. enabling traders to buy low and sell high [ahead of others].."

Based on research (pdf) by Cameron Freer and Alexander Wissner-Gross, "..many of the best places to maximize chances of buying low in one place and selling high in another (for example between New York and London) were [often] located in the world's oceans.. In the long term there is the possibility that new infrastructure deployed in new places, including on the ocean, would be advantageous. That's about all I can say at the moment, it's a very secretive industry."

It's no secret that Nova Scotia is the optimum land-based trading point between London and New York, in case your interested in making money by the speed of light. Similar to the coveted geosynchronous orbit locations in space, it's possible to imagine a future where certain land-based points on earth become high value for their shortest-distance between two or more key points (cities).
posted by stbalbach (135 comments total) 21 users marked this as a favorite
 
land-based points on earth

Should be.. land or ocean points on earth
posted by stbalbach at 7:25 PM on April 7, 2011


This is high-frequency rent-seeking behavior, in case it isn't obvious.
posted by unSane at 7:27 PM on April 7, 2011 [5 favorites]


What's worse than a bunch of dickwads wasting their education on high-frequency arbitage?

A bunch of dickwads wasting their education on high-frequency arbitage IN A FLOATING AQUAWORLD.
posted by RobotVoodooPower at 7:30 PM on April 7, 2011 [34 favorites]


Thanks, unSane. More info:

What is High Frequency trading? (WSJ)

High Frequency Trading: Wall Street's New Rent-Seeking Trick: "HFTs now represent about 70% of the trading volume in the U.S. equity market."
posted by stbalbach at 7:31 PM on April 7, 2011 [1 favorite]


it's possible to imagine a future where certain land-based points on earth become high value for their shortest-distance between two or more key points (cities).

It's possible to imagine a past in which this is so. And also a present.
posted by notyou at 7:31 PM on April 7, 2011 [11 favorites]


Well, I meant in (on?) the oceans really, since that's what the FPP is about, though I confusingly said "land".
posted by stbalbach at 7:34 PM on April 7, 2011


Furthermore, the term used to describe this behavior as practised by feudal lords who collected tolls from merchant ships travelling up and down the Rhine was Robber Baron.
posted by unSane at 7:39 PM on April 7, 2011 [5 favorites]


"It provides a new source of liquidity - or what can be viewed as a new source of liquidity - because it allows for faster turnaround times for coordinated trading of geographically separated securities," he said.

This provides nothing, except another excuse for parasites to suck yet even more of the value out of real work.
posted by PareidoliaticBoy at 7:42 PM on April 7, 2011 [8 favorites]


This is brilliant. I have this mental image of "liquidity" trying to find the level between London and NYC at Terahertz speeds.
posted by phrontist at 7:43 PM on April 7, 2011 [1 favorite]


A (large) part of me is pretty pissed off by people making money without creating any value, like in these examples. Another smaller part is sort of fascinated by the organic way the market evolves taking advantage of every possible weakness. HFT trading is like that marine dinosaur with the bizarrely long neck that allowed it to place its head in a school of prey before the prey sensed its massive body.
posted by Mei's lost sandal at 7:50 PM on April 7, 2011 [8 favorites]


Bah - high frequency trading is already a nag - bring on quantum entanglement trading!
posted by aeshnid at 8:02 PM on April 7, 2011


So, our next financial collapse based on trading questionable-worth black box financial products will be conducted at the speed of light, and only limited by the fact that light travels 186,000 miles a second.

Good to know.
posted by hippybear at 8:05 PM on April 7, 2011 [4 favorites]


I would love to see an 8 1/2 percent sales tax on stock purchases. It would put a stop to this abusive behavior and get us out of debt, and ensure that stock traders are in it for the long haul.
posted by Monkey0nCrack at 8:06 PM on April 7, 2011 [7 favorites]


The tax should be x/(t)% where t is the time in seconds that the stock is held for, and x is however long you want to set the floor for trading.
posted by Grimgrin at 8:12 PM on April 7, 2011 [7 favorites]


There's no particular reason why stock traders should be in it for the long haul. Day traders who operate without inside information do indeed contribute to the liquidity of the market, which is exremely important. The problem with this kind of HFT is that it basically constitutes a form of front running, and it's extremely sensitive to a run-for-the-exits flash crash.

All that a tax on stock trades would achieve is to drive every company to overseas exchanges.
posted by unSane at 8:13 PM on April 7, 2011 [2 favorites]


I would love to see an 8 1/2 percent sales tax on stock purchases. It would put a stop to this abusive behavior and get us out of debt, and ensure that stock traders are in it for the long haul.

That is not a good idea. There are plenty of legitimate reasons to not be "in it for the long haul" that are not detrimental to the market. Plus it would just mean the end of the United States as the center of global finance.
posted by ghharr at 8:14 PM on April 7, 2011 [2 favorites]


(If it's not obvious why liquidity is important, consider the situation where the bottom is falling out of the market. If you have a bunch of short-term traders in the market, you can probably dump your XYZ stock on the way down, since there are people still in teh market tradiing the 1-5 minute ups and downs. Without the day traders, you would be SOL and your XYZ would be going to zero and you would be stuck with it).
posted by unSane at 8:15 PM on April 7, 2011


How much liquidity is enough - or how much should be paid for?
posted by jet_silver at 8:16 PM on April 7, 2011


I'm not convinced that HFT front-running does contribute liquidity in any meaningful way. Real liquidity isn't paid for, except in terms of reducing arbitrage and increasing market efficiency. Front running is essentially a micro-tax on transactions because the guys who get in a picosecond ahead (and have the information that a trade is about to take place) can cream off a quarter cent or whatever it is on the deal.
posted by unSane at 8:19 PM on April 7, 2011


Without the day traders, you would be SOL and your XYZ would be going to zero and you would be stuck with it

Without the day traders, a single notice of "bad news" wouldn't crush the stock, thus, it wouldn't go to zero and you wouldn't need to dump it.
posted by eriko at 8:21 PM on April 7, 2011 [3 favorites]


Not previously, but previous disaster precipitated (pun!) by the same setup.
posted by Lukenlogs at 8:27 PM on April 7, 2011


Hold on, everyone, this could be the economic incentive that finally delivers an FTL drive!
posted by drdanger at 8:30 PM on April 7, 2011 [10 favorites]


What's worse than a bunch of dickwads wasting their education on high-frequency arbitage?

Well it kind of freaks me out that they're relying on computers to do it for them, and experimenting with "the physical boundaries of liquidity". With such a high volume being traded automatically based on "signs" or "signals" in the market, is it unreasonable to fear that some day this could cause a huge disaster? How many of these hedge funds make their own software? If, for example, one third of them were licensing the same or similar algorithm which could mis-perform on the same spurious signal(s), is that a major market vulnerability? There's already been one close call.

A bunch of dickwads wasting their education on high-frequency arbitage IN A FLOATING AQUAWORLD.

I'm sure there's a Bio Shock joke for this.

I guess it's no surprise that Patri Friedman and Ray Kurzweil's people are into this idea. Building floating cities filled with robot stock traders sounds like a crazy capitalist, futurist/transhumanist concept.
posted by MrFTBN at 8:31 PM on April 7, 2011


It's kind of hilarious and terrifying that Berkshire Hathaway stock movements might be correlated with Ann Hathaway being in the news.
posted by ghharr at 8:34 PM on April 7, 2011 [6 favorites]


All that a tax on stock trades would achieve is to drive every company to overseas exchanges.

You say that like it's a bad thing.
posted by PareidoliaticBoy at 8:35 PM on April 7, 2011 [2 favorites]


Without the day traders, a single notice of "bad news" wouldn't crush the stock, thus, it wouldn't go to zero and you wouldn't need to dump it.

Unfortunately, that's not true because of the existence of stop orders, which are used by many swing and long-term traders as well as day traders. Once a stock drops below a particular level (usually some round number) it triggers a wave of sell orders. Without liquidity, the price drops off the cliff, running stop after stop as it hurtles down.

Crashes happen because liquidity dries up and like it or not liquidity is provided by short term traders.

However front-running HFT is not part of this.

This is why simply taxing trades is the wrong solution. The solution is to ban front running.
posted by unSane at 8:36 PM on April 7, 2011 [3 favorites]


You say that like it's a bad thing.

It doesn't bother me as I don't live in the US but you have to decide whether you want the robber barons inside the tent pissing out or outside the tent pissing in.
posted by unSane at 8:38 PM on April 7, 2011 [2 favorites]


Can anyone explain to me why this man is wearing an electric blue fishnet suit?
posted by nebulawindphone at 8:40 PM on April 7, 2011


I don't live in the US but you have to decide whether you want the robber barons inside the tent pissing out or outside the tent pissing in.

Anyone who's been paying attention knows that the robber barons are all incontinent and are playing tag in the tent while drinking Coors Light.
posted by hippybear at 8:44 PM on April 7, 2011 [7 favorites]


they are cooling vents and placard holders, the fishnet suits.
posted by clavdivs at 8:46 PM on April 7, 2011


AquaTraders
posted by clavdivs at 8:47 PM on April 7, 2011


Can anyone explain to me why this man is wearing an electric blue fishnet suit?

Trading jackets are brightly colored and made to be well ventilated because the trading floor is active and crowded and hot. Here is a bit about them.

If you look at the guy next to him, the back of his jacket is also ventilated.
posted by hippybear at 8:48 PM on April 7, 2011


Sharkskin
posted by aeshnid at 8:49 PM on April 7, 2011


I envision a larger role in global justice for the good sea-faring folk of Somalia.
posted by Abiezer at 9:00 PM on April 7, 2011 [6 favorites]


LSE should forget the TSX and merge with these guys
posted by aeshnid at 9:01 PM on April 7, 2011


Maybe they could use some of their big piles of cash to buy more space-time bandwidth.
posted by codswallop at 9:09 PM on April 7, 2011


So when are these trading houses going to lay down their own fiber with sunken servers on the ocean bed arbitraging all day?
posted by BrotherCaine at 9:14 PM on April 7, 2011


I don't live in the US but you have to decide whether you want the robber barons inside the tent pissing out or outside the tent pissing in.

Go sell your yer half-baked LBJ quote to the rubes. That trope is about as refreshing as a drink from the Deepwater Horizon's urinals.
posted by PareidoliaticBoy at 9:22 PM on April 7, 2011


Schroedinger's portfolio?
posted by dr_dank at 9:29 PM on April 7, 2011


previously
posted by exogenous at 9:30 PM on April 7, 2011


With the amount of jobs I see advertised for algo developers, and the amount of money being poured into this area, returns have to be diminishing - there are only so many arbitrage opportunities around. Once they diminish to a certain point, one expects there will be no extra investment in the area, and the state of play at that point will roughly remain. The thing is, at that point all trading will have shifted from (inefficient) human traders to (relatively more efficient) algo traders.

In other words, equities market will have done precisely what every other area of our society is doing right now, which is rebuild its own infrastructure to take advantage of the communications revolution. This is no different from the shift from shopfront to online retailing, or from snail mail to email.
posted by claudius at 9:33 PM on April 7, 2011 [4 favorites]


What's worse than a bunch of dickwads wasting their education on high-frequency arbitage

What is with this prevailing thread on metafilter that arbitrage serves no purpose, and that traders are "dickwads"?

Do you realize that arbitrage reduces inefficiencies in the market? On an individual level, the bright people that work hard to do these jobs, in my experience, often make significant social sacrifies, both during school, and during their professional careers?

Your comment is extremely mean-spirited and rude.
posted by esprit de l'escalier at 9:43 PM on April 7, 2011 [2 favorites]


Back in my day, if you wanted to become rich, you'd do something, or work, or make a product, or sell something, or contribute SOMETHING, ANYTHING.

If there was any confusion in anyone's mind as to how our economy could be so fragile, this should clear that up. The trillions of dollars that the financial sector is worth, is based entirely on a crazy game of find-the-queen, where the the fastest card-moving algorithm wins.

Guh.
posted by braksandwich at 9:57 PM on April 7, 2011 [8 favorites]


Do you realize that arbitrage reduces inefficiencies in the market? On an individual level, the bright people that work hard to do these jobs, in my experience, often make significant social sacrifies, both during school, and during their professional careers?

OMG! Please! Won't you think of the brokers?
posted by PareidoliaticBoy at 9:58 PM on April 7, 2011 [14 favorites]


Back in my day, if you wanted to become rich, you'd do something, or work, or make a product, or sell something, or contribute SOMETHING, ANYTHING.

When, exactly, was your day? Arbitrage has been around longer than iron.
posted by pompomtom at 10:03 PM on April 7, 2011 [4 favorites]


Lightspeed arbitrage trading reduces ineffeciencies in the market. Okay. How exactly does that benefit anyone other than those doing the trading? Does it do anything for Main Street, or is it only Wall Street (and associated markets around the globe) which reap the benefits?

I've certainly seen the market doing plenty to help itself over the past 3 years, and don't see much of it making any difference in the daily life of the average person. Maybe the people who are making these vast sacrifices are doing it only for their own gain, and there's nothing about what they do which creates real value in the world.
posted by hippybear at 10:04 PM on April 7, 2011 [4 favorites]


esprit de l'escalier: The first problem I see is that HFT is conducted by brokerage houses that are acting as agents for other market participants, and the speed, volume and opacity of the decision making process behind those trades means they can mask front-running by the major banks. Essentially this is a regulatory issue.

The other problem I see, that doesn't involve fraud is what happens when the markets become as efficient as the technology can make them? This is part of what crashed Long Term Captial Management IIRC, the markets became more efficient so their arbitrage models stopped being profitable. With all the banks heavily invested in it, what if that repeats itself?

In fact, wouldn't one of the signs that HFT is actually a mask for securities fraud would be if it remains profitable even as the market becomes progressively more efficient?
posted by Grimgrin at 10:05 PM on April 7, 2011 [4 favorites]


This kind of arbitrage, and many others I can think of, is obscene and immoral. Why should someone sitting behind a screen determine the worth of someone's labor, someone they don't even know? Then to add insult to injury, that person behind a screen can actually profit by causing a farmer or other commodity worker 2000 miles away to go out of business. It's wrong; it's *damned* wrong. If there is such a thing as evil, these arbitrage activities qualify - way up near the top. Just who do these parasites think they are?
posted by Vibrissae at 10:07 PM on April 7, 2011 [2 favorites]


A bunch of dickwads wasting their education on high-frequency arbitage IN A FLOATING AQUAWORLD.

Sink that son-of-a-bitch and let the big daddies sort it out.
posted by codacorolla at 10:11 PM on April 7, 2011


Why should someone sitting behind a screen determine the worth of someone's labor, someone they don't even know?

The same reason that someone ordering a coffee determines the worth of someone else's labour? Why on earth would knowing the person make a difference?

Then to add insult to injury, that person behind a screen can actually profit by causing a farmer or other commodity worker 2000 miles away to go out of business.

The same as someone who decides that they don't like coffee that much?
posted by pompomtom at 10:12 PM on April 7, 2011 [8 favorites]


OMG! Please! Won't you think of the brokers?

You really think that childishly, sputtering contempt is something to be proud of?
posted by esprit de l'escalier at 10:13 PM on April 7, 2011


This kind of arbitrage, and many others I can think of, is obscene and immoral.

Was it also obscene and immoral when some seventeenth-century trader bought up a bunch of commodities in Village A, loaded them on a wagon, and carted them down to Village B, where he knew he could sell them at a much higher price than he bought them for? If there's a difference, what makes the kind of trading described in the FPP so much worse?

(I'm not saying there isn't a difference, but my ideas about that are only half-formed and I'm curious to hear what others might think.)
posted by twirlip at 10:16 PM on April 7, 2011 [1 favorite]


Grimgrin: That's very interesting. Honestly, I only understand the basics of why an efficient market is a good idea. I don't really understand what you're talking about...
posted by esprit de l'escalier at 10:19 PM on April 7, 2011


I used to live in Murray Hill, and it was filled with traders. I can confirm they are all, in fact, dickwads. Sorry.
posted by Threeway Handshake at 10:21 PM on April 7, 2011


I don't really understand this. Suppose we have two dealers - one in London, connected by a direct data cable to New York. The other is in the middle of the Atlantic on a boat. The dealers are waiting for a commodity to rise in price in London / New York so they can profit by selling it in New York / London. Wheat suddenly rises in price on the London market, so the first dealer sends a message to his New York counterpart. The transmission takes, say, .02 seconds.

At the time the first dealer transmits his instruction the second dealer - in the middle of the Atlantic - has not yet found out about the change in price. The second dealer discovers the change .01 seconds later, and sends his transmission to New York, taking .01 seconds and therefore arriving .02 seconds after the change in price. Result: both transmissions arrive at the same time. There is no advantage to being in the middle of the Atlantic or in fact anywhere else, as long as you're on a line connecting London to New York.
posted by Joe in Australia at 10:22 PM on April 7, 2011


Joe in Australia: there is when you're doing a zillion trades a second and each of your trades, if it arrives a fraction of a second before someone else's, can affect the trading price.
posted by hippybear at 10:25 PM on April 7, 2011


Being in the middle of the Atlantic is only to your advantage if you have access to a wired Internet connection. Otherwise you're going to be significantly delayed due to satellite lag.
posted by dunkadunc at 10:26 PM on April 7, 2011


In the telecom business, companies advertise the ultra-low latency of their routers to financial firms.

The combination of these three technologies demonstrated that tens to hundreds of microseconds can be removed from high-volume market data configurations when compared with standard software and Ethernet switching technology.


They are advertising saving 10 microseconds on stock trades.
posted by eye of newt at 10:27 PM on April 7, 2011


Pirate submarines. I'm just putting it out there.
posted by PareidoliaticBoy at 10:29 PM on April 7, 2011


Here are a few previous FPPs about HFT, which provide some interesting background and other things.
posted by hippybear at 10:32 PM on April 7, 2011


Hippybear: my point is that the time taken for anyone on a line between London and New York to
  1. Learn of an arbitrage opportunity;
  2. Issue a buy order;
  3. Issue a sell order; and
  4. Have those orders received
must be at least a minimum amount of time - the time taken for light to travel between those two points. It doesn't matter where on the line you are. In practice the most advantageous thing you can do is to be as close as possible to one end of the data cable connecting the two points, which effectively means actually being in either London or New York. As Dunkadunc pointed out, satellite connections are much slower than cable ones because the signal needs to go out into space and back.
posted by Joe in Australia at 10:35 PM on April 7, 2011


unSane: "You say that like it's a bad thing.

It doesn't bother me as I don't live in the US but you have to decide whether you want the robber barons inside the tent pissing out or outside the tent pissing in
"

I want us to singlehandedly claw our way UP that stream of piss with a hatchet in hand.
posted by symbioid at 10:37 PM on April 7, 2011 [1 favorite]


Joe in Australia: well, it would appear, according to the second link in the FPP, that some fine people from MIT have done research which directly disagrees with your position.

Not sure what to tell you there. I'm not an MIT mathematician. Perhaps you can read through the paper and find the flaw in their arguments.
posted by hippybear at 10:38 PM on April 7, 2011


Okay, assuming you aren't an expert taking the piss out of my vaguely informed post, here are the issues as I understand them.

Front running is where a broker buys a stock (or a position or a stock or some other instruments that has the effect of 'if stock X goes up person Y makes money') before placing a large order for a client. This is illegal, because in effect you're stealing from the client. By making the front run purchase for yourself, you're going to increase the price of the stock ahead of the purchase for your client, so you're increasing their cost, and possibly diminishing the profit they realize (or increasing the loss) at your own benefit.

Because the algorithms are secret, the data sources are secret and the trades are happening in microseconds there's no way to tell if these HFT algorithms are front running or not, and that's a problem. This is a separate issue from arbitrage in general.

Joe In Australia:

If I'm getting this, the missing point is that you need to know the prices in both cities before you know there's an opportunity for arbitrage.

If you're in New York, it takes you Y time to get the prices from London, and if you're in London it takes you Y time to get the prices from New York. If you're halfway between them, it only takes you y/2 to get the price from both London and New York. So theoretically you could execute the buy and sell order (assuming you have a perfect computer) at the same time that a trader in either city would become aware of the opportunity for arbitrage.
posted by Grimgrin at 10:39 PM on April 7, 2011 [2 favorites]


Was it also obscene and immoral when some seventeenth-century trader bought up a bunch of commodities in Village A, loaded them on a wagon, and carted them down to Village B, where he knew he could sell them at a much higher price than he bought them for? If there's a difference, what makes the kind of trading described in the FPP so much worse?

Thaanks for your query, and example, twirlip: I understand arbitrage, and the real market efficiencies brought by things like commercial paper, contracts, money, trading exchanges, etc. They've all been around for a long, long time; they are necessary for commerce to function.

That said, my beef is with light speed trading that combines with concomitant light speed access to information, whose access to same is created by those who are *already* wealthy. I have a very big problem with relatively new (within the last century) trading games like short selling and other more arcane kinds of arbitrage that are so unbelievably complex that even most educated people can't figure out down from up.

In short, there's a difference between adding value (like transport, in your example - a real efficiency value-add) and the kind of seeking after front end speed and advantage that ends of profiting a few and helping to enslave the many.

There are times on the crude oil commodity exchanges (for example) when 20-25+% of the cost of a barrel of crude is going to the pockets of traders. That's obscene! The same goes for other necessary commodities like foodstuffs, minerals, etc. Just look at the price of things like sugar and wheat over the last year; they have spiked in price due to *one* thing - trade exchange arbitrage. As a result, millions of poverty-stricken people, worldwide, are more challenged in their daily quest for sustenance. All this while some a**hole trader/banker and his/her bosses roll up to fancy restaurants and feed their greedy faces.

These activities need to be deconstructed, and the people who profit from those activities outed, mercilessly. Everyone on earth needs to know who these people are, and how they live; that needs to be compared with the blight that they cause. This needs to be done in a no-holds-barred manner. The arbitrage industry (and the banking industry) need a big dose of "Tunisian-style" revolution.

Going back to your 17th century example; when someone caused havoc, there was an easy way to find out who it was. Today, that's not so easy, but it's still possible. I say out the bastards (no violence!), shame them, and then change the laws that the greedy arbitrage-profiting scum pay to have passed.
posted by Vibrissae at 10:42 PM on April 7, 2011 [14 favorites]


Relevantly, there's a petition going around at the moment to get professional economists to call for a Tobin Tax:
Groups across the world are inviting economists who are qualified by post-graduate degree (Master or PhD) to sign a letter in support of a financial transactions tax (see below). The goal is 1000 signers by Friday, April 8th.
Economists can sign the open letter by entering their details in the comments box at this link: or emailing euderzo@oxfam.org.uk.

Last year, 350 economists from all over the world signed a letter in support of a financial transactions tax, and over the past year there has been significant political movement towards implementing the FTT in Europe and some other countries. The campaign for the so-called ‘Robin Hood Tax’ is now hugely popular in many countries (www.robinhoodtax.org). The French have made an FTT a priority for their presidency of the G20 and there is a real chance of a breakthrough in the coming six months.
via.
posted by wilful at 10:49 PM on April 7, 2011 [3 favorites]


Grimgrin: Yes, that should be illegal. Presumably, this could be done with some decent machine learning algorithm run by the stock exchange combined with some reporting information that a broker has to fill out after the fact. So, for example, if you get an order for a million shares of X, you record the time and date of the order, and then you fill the order using all the trades you want to do. Thirty days later, you report the order that you filled to the stock exchange, and they run some machine learning algorithm on the trades to see if you're up to anything.
posted by esprit de l'escalier at 10:52 PM on April 7, 2011


Here's a hint for knowing when a Wall Street wonk is dishing you BS -- they throw out the words "efficiency" and "liquidity" as if that is supposed to justify everything and end all conversation.

The "efficiency" these traders are adding is trivial, fractions of a cent per share. Nobody else would even notice if the high frequency traders crawled away and died. As far as "liquidity", they only supply liquidity when it isn't needed and withdraw that liquidity when it is needed most, as we saw in the flash crash. The price of that tiny amount of "efficiency" and "liquidity" is greatly increased volatility when a crisis occurs and collateral damage to innocent bystanders that far outweighs any benefits they claim. But if you talk to them, just like Lloyd Blankfein, they think they are doing God's work. Don't believe it for a second. It is just Galtian self-justification for reckless greed.
posted by JackFlash at 11:11 PM on April 7, 2011 [12 favorites]


(I'm not saying there isn't a difference, but my ideas about that are only half-formed and I'm curious to hear what others might think.)

Here's my take. The financial markets are supposed to exist to help real-world enterprises raise money, allocate risk and resources, help people invest spare capital, and so on. Without the liquidity provided by traders, it would be difficult to do this efficiently. That's awesome.

The problem is that the point we're at with this is one where the markets now exist as monsters unto themselves. In 2009, 15 trillion dollars of physical merchandise was traded between countries; every export and import combined. 17 trillion dollars of hypothetical value was traded on just the NYSE (to say nothing of London, or Tokyo or Frankfurt or...) More money was exchanged on one bank of computers than the entire world's population managed to trade in the real world.

You hear people say that if high frequency trading and other financial tinkering didn't exist, that we wouldn't be able to buy and trade stocks. Bullshit. The tail is wagging the dog. If steps were taken to reduce the financial masturbation, we could still have all the arbitrage and liquidity we needed to have efficient enough markets. If there was a 0.1% tax on every stock trade, and Microsoft is trading at a perfect objective worth of $25.00, if you wanted to buy some, someone would be thrilled to provide the liquidity to facilitate this trade at the exorbitant price of $25.03. Assuming you were investing in an actual company, which is what the markets are supposed to be for, and not trying to make a fraction of a cent of paper profit faster than you can snort a line of coke, then this would be insignificant next year when the shares are worth $28.00 or whatever. The underlying investment function of the markets, which they were providing just as well 10 years ago, (or 30 years ago for that matter) before the advent of the crazy HFT shit, is untouched.

Instead, we have thousands of very smart people spending billions of dollars trying to beat each other by 10 microseconds instead of 15. That may be the most profitable use of the capital, but it sure as shit ain't the most valuable to society.
posted by Homeboy Trouble at 11:14 PM on April 7, 2011 [9 favorites]


I say out the bastards (no violence!), shame them, and then change the laws that the greedy arbitrage-profiting scum pay to have passed.

I think maybe just a little teensy bit of violence would be okay. Pour encourager les autres.
posted by Kraftmatic Adjustable Cheese at 11:16 PM on April 7, 2011 [2 favorites]


What happened to the EU's proposed Financial Transaction Tax?
posted by Purposeful Grimace at 11:17 PM on April 7, 2011


This kind of HFT is good. I think there's a deep misunderstanding of HFT. Some is good for the market (arbitrage between different trading centers in the case here), and some are extremely bad. Arbitrage between exchanges are never evil, they ensure that prices remain the same between markets. If you buy oil futures in London, they should absolutely be the same price as the one in NY if they're the same contract, any difference means that the arbitrageur eliminates the spread and increases the likelihood that the buyer and seller pays the same price. Increasing the speed at which prices equalize can only be good. More importantly, it ensures that the liquidity pool in one market can be combined with another rather quickly, which is a VERY good thing.

There are also extremely bad forms of HFT (which isn't described in the above scenario). One kind was from something called "flash orders", which was taking information about trades BEFORE the public was able to. This flash would be sent to privledged parties (and assumed to be market makers who paid for the service) and they had the ability to fill the order themselves before the general public. This creates an unfair marketplace, obviously, that favors those who pay the exchange priveledge of access by allowing them to take better prices and be able to see changes in market direction before anyone else. Obviously, most exchanges have stopped this practice due to bad press.

Another form of HFT can that cause huge problems is from statistical time-based arbitrage from removing/providing liqudity to create a micro-squeeze. Essentally, the market isn't perfectly liquid, and an HFT shop can squeeze stock prices higher at will if certain conditions exist. This is illegal but incredibly hard (if not close to impossible) to detect. All they have to do is send a lot of buy/sell signals to the exchange if the order book looks favorable to their in-house algorithm. To put it in more simple terms, if your mutual fund wanted to buy a lot of AAPL and split their trades over the span of 30 minutes (let's say buy $100,000 worth every 5 seconds), a HFT algorithm can pick up a persistant seller and force the price up immediately before the 5-second mark and sell it back less than a second after. This is probably illegal, but the crazy thing is that some HFT shops use machine learning and the black box can be doing things like this without instructing it to do so. It's entirely possible that a hedge fund can be acting illegally without their own knowledge, because the black box trading algorithm figured out that pumping up prices before certain kinds of block trades are profitable. (Note that this example has nothing to do with the first "good" HFT example, this can all be done within one exchange.)
posted by amuseDetachment at 11:19 PM on April 7, 2011 [6 favorites]


So even if we buy the premise that arbitrage in general provides value by raising liquidity and making the market more efficient etc, does slightly faster sub-second arbitrage provide more value to the market?

If it doesn't, does slightly faster sub-second arbitrage enable more HFT skimming profit from the market, or will the faster HFT actors just take some more of the pie competing against other traders?
posted by delegeferenda at 11:30 PM on April 7, 2011


Increasing the speed at which prices equalize can only be good.

To me, this is like saying that increasing the speed your car can go can only be good. Sure, being able to go 60mph is way better than topping out at 30, and there are even occasions when 120 may be better than 60. But we're at the point where our car can do 429 mph, and there are still people spending all of their time and money tinkering with it to get it to 430. Sure, 430 is better than 429, but not in any way anyone would ever notice or appreciate. Meanwhile, all of the effort that people are spending on getting that 1 mph could be used for something else.

And the tinkering needed to get that 1 mph is also the same kind of tinkering that opens up the possibility that the car will suddenly explode, or that increases the possibility that someone can steal the car. Most of us just want the car to get us to the grocery store and don't care about that extra 1 mph. But when it explodes, everybody has to chip in to fix the damn thing.
posted by Homeboy Trouble at 11:35 PM on April 7, 2011 [6 favorites]


Increasing the speed at which prices equalize can only be good. More importantly, it ensures that the liquidity pool in one market can be combined with another rather quickly, which is a VERY good thing.

Ah, that brings up the third buzz phrase you hear from the true believers -- "price discovery."

Efficiency, liquidity, price discovery. There is no reasoned thought involved. It is just a religious chant.
posted by JackFlash at 11:36 PM on April 7, 2011


the true believers

I'm not a true believer, but I don't understand how/why the speed of transaction changes anything. I thought amuseDetachment made the best post of the thread.
posted by stbalbach at 11:40 PM on April 7, 2011


delegeferenda:
If it doesn't, does slightly faster sub-second arbitrage enable more HFT skimming profit from the market, or will the faster HFT actors just take some more of the pie competing against other traders?
It minimizes the margins to be made from arbitrage itself. By transacting faster, implicity the spread will also be tighter in normal markets (because there's less time-risk). This means buyer and seller pay a more similar price between exchanges assuming normal market conditions. The smaller time-risk (and thereby less need to be compensated for that risk) is absolutely critical to understanding what's going on here.

However, this has very interesting implications from a systemic perspective. By both operating profitably with smaller margins (due to smaller time-risk) as well as straight-up beating other parties to the punch, they are killing traditional market makers. Effectively people who have been market makers and arbitrageurs for the past 30 years are being killed off one by one because it's just not profitable to compete with automated HFT. In normal markets, you will get a more fair price. However, in less-than-normal markets (say, a market crash), many HFT shops will automatically turn their machine off in extreme events (because there's not enough data to be profitable and the risk is too high), and no one is left to fend for the markets as many of the traditional market makers are out of business or not transacting as heavily. This can create what is called a "Flash Crash". Oddly enough, the worst case scenario are those in which short-selling is not permitted or there is a very low short interest (because short-sellers are often the buyers of last resort, as they need to buy to close their positions).
posted by amuseDetachment at 11:40 PM on April 7, 2011 [2 favorites]


You can fix this without any taxes. Just batch executions into 1 second groups. Congrats, you've just made the speed of light slow enough that weird physics arbitrage ceases to be relevant. You'd need the exchanges to go along, but they're already pretty heavily regulated, aren't they?
posted by ryanrs at 11:44 PM on April 7, 2011


ryanrs: To clarify my post above (TLDR:), by doing that, the bid/ask spread (the price difference paid between a market order buyer and seller) will be bigger and people will end up paying more than they do now for their transactions. I'm fairly confident that individual small buyers will end up paying more in your example. It's not as clear as to the largest institutions (such as your mutual fund) whether they'll end up saving money from your example due to gaming of the system (although I'd argue that the gaming itself should be regulated/enforced first if possible). As recent as a decade back, stocks were priced in 1/16 or 1/32 fractions. Today, many Dow components the spread is one penny. Entire businesses exist by playing this spread, they're traditionally known as market makers or specialists. Many have shut down or switched to HFT because penny-spreads aren't profitable. It also may increase the likelihood of temporary market crashes because those traditional parties aren't as prevalent.
posted by amuseDetachment at 11:54 PM on April 7, 2011


Stock markets don't need to be in any particular place. If you're trading options (rather than the underlying shares) then the markets don't even need the share registrars to cooperate. In theory I could set up a stock market in my living room. Any attempt to impose a transaction tax or do anything else that impacts these traders' profits will immediately cause them to move to a different jurisdiction.
posted by Joe in Australia at 12:24 AM on April 8, 2011 [1 favorite]


This article got me thinking about the speed of light in different media. Are the undersea cables optimized for speed of transmission (the speed of light or signal propagation from one end to the other) or for reliability? I imagine there are undersea repeaters, and that you don't want to have to service them, and I imagine the materials in the cable (the metal or glass fiber, whatever is carrying the signal) may be optimized for long service rather than speed of signal propagation. I can imagine, naively, perhaps, someone choosing a glass fiber that is wider and less likely to break, but where the path light takes is rather long compared to a narrower fiber.

Does anyone know about the materials used in the trans-Atlantic cables, and whether they are optimized, in materials and design, for speed?
posted by zippy at 12:38 AM on April 8, 2011


The index of refraction is around 1.52, so that means light propagates at about 200,000 km/s (≅ 125,000 miles/s) through fiber optic cable. There isn't much latitude here to optimize for speed because the index must be within a certain small range of values to ensure total internal reflection -- what allows a pulse of light to travel for hundreds or thousands of km between repeaters. (And also just because there are only so many materials available that have the right properties.)
posted by Rhomboid at 12:55 AM on April 8, 2011


This reminds me of Nikolai Gogol's great novel "Dead Souls". These high frequency traders are in effect buying and selling nothing and profiting from; it's ludicrous.
posted by nikoniko at 1:08 AM on April 8, 2011


Optical fiber is engineered to maximize transparency and minimize dispersion. If you want a strong, durable optical fiber cable—and you often do—then you wrap the thing in heavy steel cable. Need more strength? Just add more steel.

The amplifiers are pretty bulletproof, too. They're just short lengths of fiber doped with a different mix of chemicals, plus a laser diode pump to supply the energy for amplification. The actual amplification occurs coherently in the doped fiber.
posted by ryanrs at 1:13 AM on April 8, 2011


They'd be singlemode fibers, right? Propagation in a waveguide isn't at the same speed as propagation in whatever medium fills the waveguide, but I really don't remember how they relate. In theory I think you could make an optical fiber with a hollow (vacuum) channel and solid "cladding", which might be a smidgen faster, if you could figure out how to manufacture it.
posted by hattifattener at 1:19 AM on April 8, 2011 [2 favorites]


Yes, single mode with many different wavelengths using the same fiber.
posted by ryanrs at 1:32 AM on April 8, 2011


Hollow fiber might violate the laws of physics, since the cladding would need to have an index of refraction less than 1. What you need is a core material with a lower index of refraction than 1.6, but still greater than 1. Magnesium fluoride comes to mind at n=1.38, but it's birefringent, which is probably intolerable. You can poke around the catalogs of weird glasses, but a lot of those probably don't have the necessary kilometer-scale clarity. Since clarity varies with wavelength, you also need to make sure you can construct optical amplifiers for the wavelengths that best suit your fibers.
posted by ryanrs at 1:50 AM on April 8, 2011 [1 favorite]


In general HFT's are exploiting the market's inelasticity. That is, prices don't always quite reflect real events, and many HFT strategies exploit the slight delays in price adjustments by attempting to detect from price patterns and other tools when they can do so. As a result they do the positive things of increased elasticity and increased liquidity.

Liquidity and in general access to capital is often underestimated as a tool for economic growth, but almost any entrepreneurial activity requires a capital outlay. When you start a lemonade stand you probably get your equipment from your parents but if you're a big boy and don't have the money? You're gonna want liquidity.

Anyways, there are also bad things. One is that it is considered general rent seeking behavior that doesn't add anything. I happen to disagree with that I value increased price elasticity and liquidity. However, due to factors such as human irrationality and imperfect trading algorithms price shocks will tend to be overamplified as the market may have become too elastic. So what would otherwise have been a slower drop suddenly turns into a crash. And a mild rise turns into a bubble in minutes.

In this particular instance this would make prices across exchanges correlate with each more closely (because opportunities for arbitrage would be snapped up).

And conversely (as usual) the speculation might not only cause smooth arbitrage that makes the price reflect true value closer, spikes of various sorts will cause the price to be WAY out of sync with actual value at time.

But is that worse than prices that are systemically maladjusted? (e.g. a situation in a related area allows the carry trade)
posted by anateus at 2:43 AM on April 8, 2011


Hollow fiber might violate the laws of physics, since the cladding would need to have an index of refraction less than 1

Less than one? How about less than zero? negative refractive index metamaterials (fascinating geek reading - not yet negative in the visible spectrum, however).
posted by zippy at 3:33 AM on April 8, 2011


If you're trading thousands of times per second, surely your major limitation is the ability to log and retain data logs? 'Cos, y'know, traders have to maintain records of every trade for several years, right? Otherwise, who is to say they're not just making shit up?
posted by scruss at 3:56 AM on April 8, 2011


I've never understood the objection to high frequency trading, arbitrage, etc. It simply doesn't fit into my definition of rent seeking.

Yes, there are large banks that cheat their own clients by giving the client the worst price out of a short window starting from the bank's real purchase. Yes, that's rent seeking, but it's already illegal insider trading too, yes? Just subpoena their source code & logs, prove your case, and fine the bastards. We aren't talking about that however.

All other purely information based human activities that need "speed limits" are fundamentally issues of civil rights. We've seen how the NSAs approach to "high frequency search warrants" is an extremely bad idea.

If you want real rent seeking, check out the commercial paper market we just blew all that money to protect. Any CEO can show growth by mortgaging hit company's real profitability for expansion.
posted by jeffburdges at 4:18 AM on April 8, 2011 [1 favorite]


I understand nothing of this. I own 100 shares in a relatively small regional bank, and I intend to let those shares sit forever, essentially. My question: is the liquidity and other advantages of this day trading and speed of light market churn really worth the parasitism it enables? 'Cos it's better be worth a lot, considering the risk and damage it produces.
posted by tommyD at 4:30 AM on April 8, 2011


tommyD: You may not personally trade shares that are affected by HFT, but your pension fund and the company that provides your life assurance almost certainly are.

jeffburges: One of the driving forces of HFT development was that you could discern the order flow of external equity investors by flooding the market with short term trades (most of which you would normally cancel within a few milliseconds) & then drop a load of orders into the book ahead of the majority of the order in question, before selling them on once you'd worked out what their limit price was. It wasn't legally front running, but it ended up having exactly the same net effect: HFT players could clip a small amount from a large number of transactions in the market at very low risk.

While there's no question that liquidity can be desirable (although that is a debatable proposition in itself) it is very hard to argue that liquidity on millisecond (or even micro-second) timescales benefits anyone except the HFT players themselves: no other player in the markets actually requires liquidity on that kind of timescale.

Inevitably, the large equity players have ended up developing their own HFT platforms in order to be able to trade without having their trades being gamed all the time by the HFT players.
posted by pharm at 4:51 AM on April 8, 2011


Any attempt to impose a transaction tax or do anything else that impacts these traders' profits will immediately cause them to move to a different jurisdiction.

That's why it would have to be global - or at least widespread enough that the costs of moving would outweigh the costs of staying where you are (let's say London introduces a tax and Bermuda doesn't; the banks won't all necessarily leave London, because there are other advantages of being there - we're seeing this at the moment with the various special taxes the UK is placing on banks - lots of talk of leaving, but nothing concrete).

We're also seeing a shift in the attitude towards a financial transactions tax. The first time I heard of it, it was being promoted by a small left-wing party in New Zealand (not saying they invented it, that's just the first time I heard of it). Now, the IMF is calling for a global FTT, as is the EU. It may not happen, but that's quite an extraordinary shift.
posted by Infinite Jest at 5:10 AM on April 8, 2011


There are times on the crude oil commodity exchanges (for example) when 20-25+% of the cost of a barrel of crude is going to the pockets of traders. That's obscene! The same goes for other necessary commodities like foodstuffs, minerals, etc. Just look at the price of things like sugar and wheat over the last year; they have spiked in price due to *one* thing - trade exchange arbitrage.

Vibrissae (or anyone else in the know), do you have a cite for this, and can you explain by what mechanism this happens?
posted by avianism at 5:22 AM on April 8, 2011 [1 favorite]


Is Broker-hate the new anti-intellectualism?
posted by blue_beetle at 5:55 AM on April 8, 2011 [3 favorites]


Are you describing an inherent feature of any exchange like system that provides certain necessary properties pharm, ala order and cancel. Or just some bug in our current stock exchanges?

In fact, you could obviously have exchanges that didn't support cancelation of orders, only expiration. I'd imagine that'd halt the particular front running you describe. Is such an exchange undesirable to institutional investors for some reason?

If it's a bug, we should fix it. If this warrants redesigning the exchanges, that's fine too. Yet otherwise, we best take reality into account and avoid trying to legislate against physics.

You might phrase this as "Inherent Cat is Inherent" or take this wonderful quote that I'd usually deploy in support of piracy :

The very powerful and the very stupid have one thing in common.
Instead of altering their views to fit the facts, they alter the facts
to fit their views ... which can be very uncomfortable if you happen to
be one of the facts that needs altering.
-- Doctor Who, "Face of Evil"

There are many people being made nervous by bitcoin too, but if its approach is fundamentally viable and it eventually eliminates distributes some powers of government to anyone who buys the computing power, well that's wonderful.

In particular, I'd imagine that an asynchronous exchange that properly constrains the underlying trading operations, ala eliminating order cancellation, might be preferable to the synchronous systems being proposed up thread, ala orders are processed only once per second.
posted by jeffburdges at 6:17 AM on April 8, 2011


Anti-intellectualism is generally a bad thing, as such it differs from 'broker-hate'.
posted by asok at 6:25 AM on April 8, 2011 [1 favorite]


ugh - there is no more overyped issue then HFT. The fact that being co-located is now paramount to any strategy (and the NS thing is related to that) tells you how commoditized it has already become. The only people it rips off are speculators not investors, is not any sort of systemic risk given the position holding times, and frankly with the SEC being defunded by the second I would much rather see them focus their time and money on other more pressing issues.

As someone who does buy and hold investing for a living (our average holding time is more than 2 years at this point, and getting higher by the day), I am happy to let the HFT guys beat the hell out of the day traders, churning mutual fund managers and one another while I just sit back. And on those rare days the algos freak out I am happy to pick up shares on the cheap.
posted by JPD at 6:27 AM on April 8, 2011 [4 favorites]


I mean think about it - how good a business is something when your success is determined by how many feet away you are from the exchanges server.
posted by JPD at 6:29 AM on April 8, 2011


Where anti-intellectualism = distrust of intellectuals
and broker hate = distrust of brokers
posted by asok at 6:53 AM on April 8, 2011


the funny thing is that HFT would be more pissed at you calling them brokers then at them hating you.
posted by JPD at 6:58 AM on April 8, 2011 [4 favorites]


fuck. You hating them.
posted by JPD at 6:58 AM on April 8, 2011


A (large) part of me is pretty pissed off by people making money without creating any value, like in these examples.

There are plenty of stupid, asinine things people do that create boatloads of wealth and nothing of value. Take Lady Gaga! (/ducks). That's not my problem with it, not in the slightest.

My problem is the complete lack of fairness in access, which inevitably leads to cronyism. Who gets premier access to the lines coming out of the Atlantic (hint: not you), and who are the unwashed, unprivileged masses that have to piggy-back off of those favored interests?

Ever try play an online game against a competitor with zero-lag while you're stuck behind a dial-up modem? You'll get slaughtered, time and time again.

Same problem.
posted by Civil_Disobedient at 7:08 AM on April 8, 2011 [2 favorites]


Actually C_D access isn't really an issue. The exchanges all have pretty clearly written policies on the issue. Its part of why the business has commoditized so quickly.
posted by JPD at 7:10 AM on April 8, 2011


jeffburges: Are you describing an inherent feature of any exchange like system that provides certain necessary properties pharm, ala order and cancel. Or just some bug in our current stock exchanges?

It's not a bug, it's a feature! Fundamentally, as I understand things, if an exchange allows you to cancel arbitrary orders in the queue then you can play all sorts of games on very short timescales.

In fact, you could obviously have exchanges that didn't support cancelation of orders, only expiration. I'd imagine that'd halt the particular front running you describe. Is such an exchange undesirable to institutional investors for some reason?

Institutional investors have to go where the liquidity is. If the exchanges all offer access to HFT traders as well, then they don't have a lot of choice. Some investment banks offer 'dark pool' exchanges which I believe some institutional investors use, as well as implementing their own HFT platforms.

If it's a bug, we should fix it. If this warrants redesigning the exchanges, that's fine too.

Part of the problem is that the institutions with the market clout to setup separate trading platforms that might compete with the existing exchanges are generally already profiting significantly from HFT, so they have no interest in fixing the system. Likewise the exchanges themselves are private businesses that profit from the increased order flow. There's a lot of profit being made in HFT, and that profit buys political access, so I wouldn't expect anything to change any time soon.
posted by pharm at 7:44 AM on April 8, 2011 [1 favorite]


Civil_disobedient: AFAIK, access to server hardware in a rack with a low-latency connection to an exchange (or even in the exchange itself) is available to pretty much anyone with the cash. Rent a server on Euronext & off you go...

There are a pile of algorithmic trading houses around who'll happily take all your money off you if you don't know what you're doing though.
posted by pharm at 7:51 AM on April 8, 2011


pharm - I think part of the issue as well in the low return on capital for a new entrant today. WRT to fixing HFT and the exchanges. Actually I'm of the opinion that Exchanges themselves are pretty much doomed unless there is some sort of government intervention. I'm also not totally sure that I care. I suspect if dark pools have issues it will work in the favor of people happy to see more volatility like myself.
posted by JPD at 7:58 AM on April 8, 2011


The exchanges all have pretty clearly written policies on the issue.

Where?
posted by Civil_Disobedient at 8:09 AM on April 8, 2011


click on Pharm's link. It tells you how to rent co-located space. That's the only barrier to entry
posted by JPD at 8:18 AM on April 8, 2011


Wow, all this effort and energy dedicated to not creating anything of value. Will our grandchildren forgive us?
posted by Afroblanco at 8:25 AM on April 8, 2011 [1 favorite]


The "efficiency" these traders are adding is trivial, fractions of a cent per share. Nobody else would even notice if the high frequency traders crawled away and died.

If this were true, it wouldn't be so profitable.

There are times on the crude oil commodity exchanges (for example) when 20-25+% of the cost of a barrel of crude is going to the pockets of traders. That's obscene! The same goes for other necessary commodities like foodstuffs, minerals, etc. Just look at the price of things like sugar and wheat over the last year; they have spiked in price due to *one* thing - trade exchange arbitrage.

This is just ignorance masquerading as outrage. Instead of enacting your plan of "shaming the traders," why not take an economics course?

The value that is being provided by an arbitrageur in this case is decreased volatility in crude oil. It may not be significant to you whether you have to pay $2.50 per gallon of gas one week, or $5.00 the next. But, that's because you have savings and can essentially act as your own arbitrageur. However, if you're a running a trucking company that buys hundreds of thousands of gallons every day, then the difference can easily overrun your cash reserves in one day, and drive you out of business in a matter of weeks. Arbitrage makes business possible by reducing volatility. That's the created value.

The idea that you should decide what fraction of every price tag of a commodity belongs to every player in the market (5% to the farmer, 20% for shipping, 40% for the supermarket, etc.) is the height of arrogance. Who are you to decide the value of anyone's work?

all this effort and energy dedicated to not creating anything of value.

You mean, your comment? ;)
posted by esprit de l'escalier at 9:10 AM on April 8, 2011




esprit, so you're saying that these high frequency traders have decreased (are decreasing) volatility in the markets? Bollocks! Markets have been acting extremely manic; if anything, these high frequency traders seem to have figured out a way of profiting off of (and feeding) this increased market volatility.
posted by nikoniko at 10:58 AM on April 8, 2011 [1 favorite]


"profiting off of (and feeding) this increased market volatility" .... by decreasing the irrational volatility.
posted by esprit de l'escalier at 11:35 AM on April 8, 2011


This is like the Galileo affair: a bunch of people who are suspicious of what they don't understand, irrationally clinging to unfounded beliefs.

Imagine a store owner who buys eggs every week. If there's volatility in the price of eggs, he might buy extra when the eggs are cheap. But, to do that, he has to have extra capital, and he has to guess how many eggs he's going to sell. If he guesses high, the eggs go bad, but if he guesses low, he can't make as great a use out of the volatility. Thus, the better his guesses, the more money he can save by buying low. Let's he hires someone to guess how many eggs he can sell. That person is called an arbitrageur.
posted by esprit de l'escalier at 11:41 AM on April 8, 2011


huh? That's not really what's going on here. Its more like "Store owner needs to buy eggs. Egg farmer is willing to sell him eggs for .95. "Arbitrageur" actually hears store owner will really pay up to a dollar. Tells Egg Farmer he'll buy all his eggs for .96 cents. "Arbitrageur" then sells his eggs to the store owner for .97"
I mean that vastly over inflates the scale of it, but that's really what's going on. It doesn't really bother me, because I as the store owner think I'm going to able to sell my eggs for a much much bigger profit then the "arbitrageur" made, but still.
posted by JPD at 12:30 PM on April 8, 2011


I'm sorry, but that's completely wrong. Why wouldn't the store owner buy the eggs directly?
posted by esprit de l'escalier at 12:49 PM on April 8, 2011


There are times on the crude oil commodity exchanges (for example) when 20-25+% of the cost of a barrel of crude is going to the pockets of traders

I'm interested in this topic, but have trouble following it. espirit de l'escalier, maybe you can take a shot at answering a few questions which still seem open to me—first, is it true that at some point up to 20-25+% of the cost of crude has gone to traders? Where would one go to get a source for data like this?

And, regardless of the truth of the claim above, two further questions. 1) How does a portion of the money paid for e.g. crude (or eggs, or whatever) get diverted to traders or arbitrageurs, exactly, and 2) how is the amount of money diverted determined?
posted by avianism at 1:08 PM on April 8, 2011


To extend the analogy further to make it clear, esprit de l'escalier, the store owner wouldn't buy the eggs directly because in this situation, the buyer and seller are far away from each other in differen markets.

Let's say 10 years ago, the trading price was 0.95 for buying and 1.00 for selling in the seller's local market. There was a distributor that would always buy it for 0.95 and sell it to anyone, anywhere, anytime for 1.00 and net a 0.05 margin. This distrubtor was always around to buy and sell eggs, but would only buy/sell at a 0.05 spread (in this situation this distributor is your traditional old-school market maker and made money off this spread).

This new Hen Factory Trader, or HFT, comes along and notices that the distributor is making way too much money off the backs of farmers and supermarkets and decides to cut out the distributor. They are willing to profitably buy and sell eggs with low margins using robot drivers and automated logistics. The HFT feels a little sketchy because a robot call the farm and supermarket every day to arrange a price with a 0.01 spread. The farm and supermarket are happier and make more money because the farm sells for 0.96 and the supermarket buys for 0.97, a more favorable price for both parties. The robot transport makes sure they can transport the eggs really quickly, reducing the chances for spoilage and eliminate any time-based price risk while transporting the eggs (if the prices go down a lot while delivering the eggs, the HFT loses money). This rapid delivery of eggs makes sure that they can profitably deliver with lower margins (with less cost passed on) between markets. The risk, however, exists because the HFT is a bit of a fly-by-night operation, and is undercutting the traditional distributors so they're going out of business. If the HFT doesn't make their call to the egg farmer and supermarket to match a price, then things break down really fast for all parties. However, if this doesn't happen and things remain stable, the farmer and the supermarket save a LOT of money in the meantime and the old-school distributor is the only one losing out.
posted by amuseDetachment at 1:09 PM on April 8, 2011


(And, another question: is the notion of money being "diverted" even the right way of framing those previous questions? It seems like in almost every discussion about this topic, almost everybody involved is taking some answer to this particular question for granted, whereas to me it seems pretty unclear.)
posted by avianism at 1:13 PM on April 8, 2011


avianism:

Commodities futures traders are also misunderstood, like HFT. I think the only trading activity more misunderstood than them are short sellers (in my opinion, intelligent short sellers that do deep research are the most honorable of all market participants).

It's definitely possible that 25% of market activity is paid to traders, you can see it a lot in periods where there is a shift towards more extreme contango. E.g. if a commodity was priced at $100 today and a contract to lock in delivery of the commodity one year from now costs $110 today, however some market shocks (say, a hurricane) makes delivery more difficult and the cost delivery one year from now changes to $150 today. Any commodity trader that locked in a delivery for $110 one year from now makes a ton of money. The market is saying one year from now, this commodity will be really rare and the world doesn't have enough supply of that commodity.

There will be several reactions to a low price of delivery today (spot price), compared to a high lock-in contract for delivery price one year from now due to an external shock (like the aforementioned hurricane). They are all "good" for the market. The price of the commodity for delivery will start to stabilize by raising the cost of the commodity today and lowering the cost of delivery one year from now. What's going on and why does this happen? Hedge funds will start to buy commodities for delivery today (spot) and store them in warehouses and sell contracts for delivery one year from now. This is "good" for the market because the commodities market is saying that there will be a huge shortage of something in the future and if we have too much of something today, we can plan for it for the future. Secondly, the actual miners/farmers would try to increase production of a commodity. An example of this are the Alberta Oil Sands. When oil is above ~$80/barrell, they can profitably extract oil. So whenever future prices of oil goes above that price, the oil sands company will lock in delivery and start producing, which prevents oil prices from rising even further.

This is how a free market plans for how many barrels of oil to extract from the ground every year and how many pigs we raise. This is how we know how much to make of anything, how much to save for the future. If we don't have these profit incentives, people wouldn't bother to store goods for the future or try to increase to supply of commodities in the future. I suspect the quoted 25% margin is from the speculator that predicted that oil prices will rise in the future and bought it today. This is useful for the markets, because it encourages oil producers to figure out ways to find oil to sell for delivery in the future today. If the speculator is wrong, they'll lose out big one year from now when we make way too much oil and they own a bunch of overpriced delivery contracts.

The time to be suspicious for manipulation of markets and making crazy money off commodities is when a commodity is normally in contango and it goes into extreme backwardation. That could be (but definitely not always) some kind of market manipulation to buy out all supply of a good to squeeze out the front-month contracts.
posted by amuseDetachment at 1:33 PM on April 8, 2011 [1 favorite]


Also, here is a TEDx talk on algorithmic trading that may be tangentially relevant.
posted by amuseDetachment at 1:47 PM on April 8, 2011


to add to aD's excellent follow-on - the reason why the store can't go straight to the farmer isn't really distance - most of the time it is because our mythological world has weird hens that will not lay reliably - so if you didn't use some distributor you would have to call every single hen house you knew until you filled your order. Historically you were happy to pay the distributor a nickel for doing this. And you had a different distributor for bananas and another for bread.

So aD and I have painted electronic market making as some purely benevolent thing, and that's really not the case - not only do HFT pull liquidity out when its needed most, but they do all other sorts of weird things like fake bids to figure out if our hen house will really sell at .95 and our market will really only buy at up to a dollar. They really want to use those fake bids to convince the hen house to sell at .94 and the store to buy at 1.01 if they can.

Also calling some commodity trader who made money on a position as capturing some sort of rent is false. Economically its the same thing as buying a stock that went up.
posted by JPD at 3:12 PM on April 8, 2011 [2 favorites]


Wow, all this effort and energy dedicated to not creating anything of value.

Everything that I see being decried here is part and parcel with private property and capitalism. It would be ludicrous to try to operate society along those lines without activities like arbitrage. I think the issue is obscured a lot by these acronyms and confused ideas people have about what algo-traders actually do.

A good example of arbitrage are these guys who trawl used book stores with database-polling, barcode-scanner equipped, PDAs. I live in northern Virginia, enjoy browsing used book sales/shops (though I rarely buy anything), and see them constantly.

They scan hundreds of books every day and applications on their phones check how much that title has sold for lately on various online market places. Presumably they have some formula to figure out whether a purchase would be worth their while.

So these guys buy a lot of books that might otherwise molder on the shelf, they get paid something for it, and I, as a frequent buyer of used books on Amazon/Ebay/Abebooks get access to all sorts of obscure titles for way less than otherwise. It's that last part everyone here seems to be ignoring.

These market-playing algo dudes are making it cheaper for someone to buy something, concomitantly increasing sales the seller's sales, though admittedly the "products" involved get extremely abstract (e.g. the right to indirectly vote in corporate governance and extract dividends from a pool of software companies).

This kind of arbitrage, and many others I can think of, is obscene and immoral. Why should someone sitting behind a screen determine the worth of someone's labor, someone they don't even know?

Capitalism just can not work effectively for anyone without this kind of activity. I'm on the libertain socialist end of the political spectra, and so agree that the effects of this kind of thing can be tragic, but trying to mitigate the excesses of capitalism by arbitrarily deeming some sorts of arbitrage off limits seems futile and unpredictably harmful.

Algotraders are just very clever players in a screwed up game, and contribute equally to it's benefits and drawbacks.
posted by phrontist at 4:01 PM on April 8, 2011 [2 favorites]


I said: This kind of arbitrage, and many others I can think of, is obscene and immoral. Why should someone sitting behind a screen determine the worth of someone's labor, someone they don't even know?

someone said (my emphasis):
Capitalism just can not work effectively for anyone without this kind of activity. I'm on the libertain socialist end of the political spectra, and so agree that the effects of this kind of thing can be tragic, but trying to mitigate the excesses of capitalism by arbitrarily deeming some sorts of arbitrage off limits seems futile and unpredictably harmful.


Capitalism doesn't work for everyone - not by a long shot; not the way we have designed it. Buy a round-the-world airfare and hang out for a while in Africa, South America, Southeast Asia, the Indian subcontinent; Central America, and you might change your philosophy.

Also, someone above asked for a cite to my earlier statement about traders often taking huge profit percentages from the price of a commodity. I should clarify and say that probably 25% of the cost of a commodity is due to the activity of traders. I was wrong to state that traders take 25% of the price; they don't. They *cause* the instability. They are *mostly* the reason for shortages, high prices, etc. In fact, traders essentially *make* markets, they don't "participate" in them, per se. Ironically, they are *enabled* by our so-called "free-market'" system, which is free-market in name, only. How, for instance, is a market free, when government subsidies are permitted to prop up one country's commodity prices in ways that let that country A's commodity producers undersell country B's commodity producers. Look at corn wipeouts by large American agribusiness groups, following the implementation the US Farm Bill, some years ago. What was "free market" about that? Other examples abound.

One example: take a look at what the Hunt brothers did with silver some years ago. They used arbitrage to completely dominate a market. Also, look at food prices, even though many food commodities have had boom production years. Prices are climbing because traders are playing the markets with all kinds of games, and they enrich themselves *at the expense of others*.

I'm a capitalist, and I believe in capitalism's future (as we evolve toward more sharing of wealth because it *makes sense*). However, capitalism has no future if this sort of privileged access to, and trading on raw information, purely for the enrichment of few, continues. Things have got to change; we need to *remove* bad actors from the system, legally.

This is not about redistributing wealth; it's about making sure those who are the source of wealth actually gain some of it back, in the end - instead of being manipulated and bankrupted by greedball traders.
posted by Vibrissae at 4:53 PM on April 8, 2011 [1 favorite]


Vibrissae: I think many of your examples are of commodities markets not being transparent and I am entirely for fair markets which pay the farmers and actual producers well without any unfair privilege towards first-world nations.

However, I would contend with the idea that traders cause widespread negative instability in the commodities markets. A lot of the instability you see is due to the futures curve (prices on delivery of a commodity across different points in time). For example, the shock with oil prices were caused by traders bidding up the price of spot oil, which increases the cost of your gas at the pump today. Traders were 100% in control of that, but you really have to look at why they were bidding it up and whether it was ultimately bad. The reason traders bidded up oil a couple years ago to record high was due to peak oil. Sometime in the past five years we have reached peak oil in terms of cost -- the cost of producing one barrel of oil will be forever rising for the foreseeable future. It was at that point a good deal to store all the oil you can find for future years. Yes the traders are making money and controlling and raising the price of oil, however they were doing it because the cost of oil is rising and it's better to use less today (raise the cost today) because we won't have enough tomorrow (increase supply and lower cost tomorrow). If they're wrong, they're stuck with a bunch of cheap oil. The Hunt brothers saw the end result of being wrong about future supply/costs with silver, they lost a ton of money. Being wrong means you lose a lot of money (the money goes to both speculators who take the opposite position and the people who actually produce the good).

We're seeing similar things with food today, especially more wealthy food such as meats. There is a big trade betting on China and India becoming more wealthy and bidding up the cost of high-class foods, you see some hedge funds making some big bets on this with things like Cocoa. By raising the price of delivery of foods today, it encourages farmers to make more food for tomorrow. This obviously has an impact for food security for all nations because food prices are rising for other nations that cannot afford it, but the cause of this isn't traders manevolently manipulating things to screw over your average 3rd world citizen. It's traders reacting to 2 billion people getting out of poverty. Less developed economies are facing the fallout, but ultimately higher food futures prices means more food gets made and farmers make more money. I don't know the solution to this, but I do know that restricting speculators from bidding up the prices of food means lowering the amount of food that will be made one year from now. This is also why corn ethanol is a ridiculous proposition socially at this point in time.
posted by amuseDetachment at 5:13 PM on April 8, 2011


Vibrissae: Who said it worked for anyone? I didn't.
posted by phrontist at 5:14 PM on April 8, 2011


".....you see some hedge funds making some big bets on this with things like Cocoa. By raising the price of delivery of foods today, it encourages farmers to make more food for tomorrow. This obviously has an impact for food security for all nations because food prices are rising for other nations that cannot afford it, but the cause of this isn't traders manevolently manipulating things to screw over your average 3rd world citizen. It's traders reacting to 2 billion people getting out of poverty. Less developed economies are facing the fallout, but ultimately higher food futures prices means more food gets made and farmers make more money. I don't know the solution to this, but I do know that restricting speculators from bidding up the prices of food means lowering the amount of food that will be made one year from now. This is also why corn ethanol is a ridiculous proposition socially at this point in time."

Good points. However, it continues to irk me that traders - people with position, access, power, and (often) wealth, profit far more from their efforts to "secure the commodity future" than those who produce the commodities. I don't have an answer, other than pointing it out, and perhaps inspiring others to point it out. I just can;t stand the fact that a trader is riding in a Rolls because s.he made a good hedge on the labor of some poor cocoa farmers in Ivory Coast. Damn, that's just not right - nor is it right when commodity priced get bid down. It's just not right.
posted by Vibrissae at 10:13 PM on April 8, 2011


amuseDetachment: One more thing, the link I put up above is an eye opener. Here's an excerpt:

"The impacts of commodity speculation:

"The increase in the price of food has been disastrous for people across the world. There were 75 million more hungry people in 2007 and a further 40 million in 2008.[7] The latest estimate by FAO in June 2009 was that over one billion people are now chronically malnourished due to 'the global economic slowdown combined with stubbornly high food prices'.[8]

"But the impact of high prices goes well beyond not getting enough to eat. Poor households in Southern countries tend to spend between 50% and 90% of their income on food, compared to an average of 10-15% in Northern countries.[9] It is estimated that the food price spike increased the number living in poverty by between 100 and 200 million.[10] As well as eating less food, households have been forced to:

* Eat less fruit, vegetables, dairy products and meat in order to afford staple foods.

* Reduce any savings, sell assets or take out loans.

* Reduce spending on 'luxuries' such as healthcare, education or family planning.[11]

"Women tend to manage the food budget and often bear much of the suffering. Women may also try to increase income through taking on insecure and risky employment such as becoming domestic workers, mail-order brides and sex workers.[12]

"High food prices affect poor farmers as well as the urban poor. A high percentage of rural households are net buyers of staple foods. In Kenya and Mozambique, around 60% of rural householders are net buyers of maize.[13] Very few poor farmers produce a significant surplus to sell.[14] In Zambia, 80% of farm households grow maize, but fewer than 30% sell any.[15] In addition, any increase in income was for many producers negated by increasing costs of farm inputs such as oil and fertiliser. The cost of fertiliser almost doubled in 2007 and 2008.[16]

"Furthermore, in general terms wild price swings make it difficult for farmers to make decisions about what crops to grow and in what they should invest precious resources. As Prof. Jayati Ghosh says: 'The world trade market in food has started behaving like any other financial market: it's full of information asymmetry . So farmers think, "Well, wow, the price of sugarcane is really high," and they go out there and cultivate lots of sugarcane. By the time their crop is harvested, the price has collapsed. So you get all kinds of misleading price signals. Farmers don't gain.'[17] "

..end excerpt...
posted by Vibrissae at 10:41 PM on April 8, 2011 [2 favorites]


amuseDetachment and phrontist: I agree.

Vibrissae: You're talking about the increase in food prices caused by commodity speculation? The advantage to increased food prices is that farmers get paid more. That's great because it means that they will produce more food, which will then keep food prices down. The solution to high food prices causing starving people is increasing minimum wages and social safety nets in developing countries. As human beings, we need this to happen, but I don't know how we can change laws in other countries. We would probably need to do it on a global level because any lone developing country that increases its minimum wage destroys its industry (I imagine.)

Vilifying arbitrage because it raises the price of necessary commodities doesn't make sense to me.
posted by esprit de l'escalier at 10:56 PM on April 8, 2011


Grimgrin: "The tax should be x/(t)% where t is the time in seconds that the stock is held for, and x is however long you want to set the floor for trading."

Brilliant! You have my vote.
posted by InsertNiftyNameHere at 1:49 AM on April 9, 2011


Vibrasse:
I just can;t stand the fact that a trader is riding in a Rolls because s.he made a good hedge on the labor of some poor cocoa farmers in Ivory Coast. Damn, that's just not right - nor is it right when commodity priced get bid down. It's just not right.
I'm not sure how to say this without sounding a little bit callous, but the global impact of an average farmer screwing up isn't that big. The global impact for a commodity trader screwing up is massive. People will starve or we will have to throw billions of dollars of grains away due to spoilage if the commodity traders doesn't accurately predict grain prices/production one year from now. This is serious business and they have to be compensated for it. Do you have a problem with doctors or engineers getting paid generously for performing complex work as well?

If you think people are starving because of commodity traders, you're dead wrong. People are starving in Africa because people in India and China are eating more and willing to pay farmers more for their food. It's that simple. Hypothetically, if grain prices rise one year then stay stable at those higher prices for the following years, the commodity speculator only makes money that first year, every year after pretty much all the money goes directly from the consumer to the farmer. Commodity speculators make money off shifts in demand/supply due to changing demographics, extreme events, etc. only, as it should be. They make money when they send the sigal to farmers, "hey you need to grow more/less rice" and are correct. If they screw up, people will starve, but I see it as a better idea to make sure that you have a group of people that know exactly how much rice to produce in a given year.

You also have to understand, this isn't some kind of magical fairlyland where we can grow as much rice/corn/etc. as possible and feed everyone guaranteed 100% with no implications. Food production has a cost and farmers must be paid. If we decreed that rice is at $5, a lot of farmers cannot make money from rice that low (they have to pay for labor, water, fertilizer, etc), so production goes down. If we decreed that rice was at $20, a lot more will be made, but people will not be able to afford it. The market is sending a signal that Indians and Chinese want more food and are willing to pay for it. The commodity speculator has to balance these factors and to ensure that farmers get paid and people will actually buy the stuff that's produced at that cost. This isn't some kind of evil conspiracy to screw the 3rd world out of their money, some are just getting screwed because the average chinese citizen wants more food, as their income has gone up by an order of magnitude. They're willing to pay the farmer that has crap land to lay on some fertilizer so they can eat more rice.

Yes it's kind of lame that farmers will never be able to see the money that commodity speculators do, but they will also never see the money that a professor of engineering that does consulting on the side makes. They do things that can seriously impact the world and are compensated for not screwing up -- the solution is not removing the incentive for getting the right amount of food to make next year.
posted by amuseDetachment at 3:12 AM on April 9, 2011


Eh - the sort of commodity trader who is making sure Korea has enough wheat at a reasonable price is a very different person from chocfinger rolling around in his Virage
posted by JPD at 6:51 AM on April 9, 2011


Obviously I'm doing it wrong. I buy a stock because I have a reasonable expectation that the company will perform well in the coming decade and I hold it for about that period of time, barring some radical event.
posted by dgran at 11:09 AM on April 11, 2011


« Older David Armand explores the medium of interpretative...   |   A specialized type of planetary gear. Newer »


This thread has been archived and is closed to new comments