User: Password:
|
|
Subscribe / Log in / New account

Linux at NASDAQ OMX

LWN.net needs you!

Without subscribers, LWN would simply not exist. Please consider signing up for a subscription and helping to keep LWN publishing

By Jonathan Corbet
October 21, 2010
One tends to think of "the NASDAQ" as a single exchange based in the US, but, in fact, NASDAQ OMX operates exchanges all over the world - and they run on Linux. In the US for instance, that includes markets like the NASDAQ Stock Market, The NASDAQ Options Market, and NASDAQ OMX PSX, its newest market that launched on October 8. At a brief presentation at the Linux Foundation's invitation-only End User Summit in Jersey City, NASDAQ OMX vice president Bob Evans talked about the ups and downs of using Linux in a seriously mission-critical environment.

NASDAQ OMX's exchanges run on thousands of Linux-based servers. These servers handle realtime transaction processing, monitoring, and development as well. The big challenge in this environment, of course, is performance; real money depends on whether the exchange can keep up with the order stream. Latency matters as much as throughput, though; orders must be responded to (and executed) within bounded period of time. Needless to say, reliability is also crucially important; down time is not well received, to say the least.

To meet these requirements, NASDAQ OMX runs large clusters of thousands of machines. These clusters can process hundreds of millions of orders per day - up to one million orders per second - with 250µs latency.

According to Bob, Linux has incorporated some useful technologies in recent years. The NAPI interrupt mitigation technique for network drivers has, on its own, freed up about 1/3 of the available CPU time for other work. The epoll system call cuts out much of the per-call overhead, taking 33µs off of the latency in one benchmark. Handling clock_gettime() in user space via the VDSO page cuts almost another 60ns. Bob was also quite pleased with how the Linux page cache works; it is effective enough, he says, to eliminate the need to use asynchronous I/O, simplifying the code considerably.

On the other hand, there are some things which have not worked out as well for them. These include I/O signals; they are complex to program with and, if things get busy, the signal queue can overflow. The user-space libaio asynchronous I/O (AIO) implementation is thread-based; it scales poorly, he says, and does not integrate well with epoll. Kernel-based asynchronous I/O, instead, lacks proper socket support. He also mentioned the recvmsg() system call, which requires a call into the kernel for every incoming packet.

There is some new stuff coming along which shows some promise. The new recvmmsg() system call can receive multiple packets with a single call. For now, though, it is just a wrapper around the internal recvmsg() implementation and does not hold the socket lock across the entire operation. But, he said, recvmmsg() is a good example of how the ability to add new APIs to Linux is a good thing. He also likes the combination of kernel-based AIO and the eventfd() system call; that makes it possible to integrate file-based AIO into an applications normal event-processing loop. There is also some potential in syslets, which he sees as a way of delivering cheap notifications to user space; it's not clear whether syslets will scale usefully, though.

What NASDAQ OMX would really like to see in Linux now is good socket-based AIO. That would make it possible to replace epoll/recvmsg/sendmsg sequences with fewer system calls. Even better would be if the kernel could provide notifications for multiple events at a time. Best would be if the interface to this functionality were completely based on sockets. He described a vision of an "epoll-like kernel object" which would handle in-kernel network traffic processing. The application could post asynchronous send and receive requests to the queue, and receive notifications when they have been executed. He would like to see multiple sockets attached to a single object, and a file descriptor suitable for passing to poll() for notifications. With a setup like that, it should be possible to push more network traffic through the kernel with lower latencies.

In summary, NASDAQ OMX seems to be happy with its use of Linux. They also seem to like to go with current software - the exchange is currently rolling out 2.6.35.3 kernels. "Emerging APIs" are helping operations like NASDAQ OMX realize real-world performance gains in areas that matter. Linux, Bob says, is one of the few systems that are willing to introduce new APIs just for performance reasons. That is an interesting point of view to contrast with Linus Torvalds's often-stated claim that nobody uses Linux-specific APIs; it seems that there are users, they just tend to be relatively well hidden.


(Log in to post comments)

Linux at NASDAQ OMX

Posted Oct 21, 2010 17:31 UTC (Thu) by intgr (subscriber, #39733) [Link]

He described a vision of an "epoll-like kernel object" which would handle in-kernel network traffic processing. The application could post asynchronous send and receive requests to the queue, and receive notifications when they have been executed. He would like to see multiple sockets attached to a single object, and a file descriptor suitable for passing to poll() for notifications.

Sounds a lot like the kevent notification interface

to contrast with Linus Torvalds's often-stated claim that nobody uses Linux-specific APIs.

It always bothers me that free software operating systems aren't cooperating to draft out new APIs that everyone could use. It might not work out in the end, but I'd like to see people at least try.

Linux at NASDAQ OMX

Posted Oct 21, 2010 18:44 UTC (Thu) by ceswiedler (guest, #24638) [Link]

I think it sounds lot like Windows IO completion ports--something the Win32 got right. More recent kernels can dequeue multiple events in a single system call.

Linux at NASDAQ OMX

Posted Oct 21, 2010 19:03 UTC (Thu) by bcrl (subscriber, #5934) [Link]

The kernel's AIO APIs were designed to allow for the pulling of events off the completion queue in userspace, which would be very efficient. I never got a chance to implement that owing to doing other stuff for work in recent years, but if there's a will, there's definitely a way to improve things without introducing yet another set of APIs.

Linux at NASDAQ OMX

Posted Oct 21, 2010 22:26 UTC (Thu) by Lennie (guest, #49641) [Link]

I think Linus should look better, lots of software uses a Linux specific API, mostly on the performance side though.

But I can definitely understand why adding an other Linux specific API might be a bad idea.

I fully agree, we could really use new API which everyone will use.

Or adapt something like kevent.

The problem might be everyone needs to agree on the API and the API should get the most performance on every platform. And I wouldn't be surprised if many of the implementers do not agree on the way to go about this.

Linux at NASDAQ OMX

Posted Oct 21, 2010 23:24 UTC (Thu) by dlang (subscriber, #313) [Link]

most software authors are still trying to make their code portable across different flavors of *nix. those programmers aren't going to use Linux specific APIs.

there are some people who are interested enough in performance to make their software linux-only, or maintain the ifdef variations (and testing overhead) of using different code on linux than on other systems.

overall, linux-specific APIs get very little use, but usually the people advocating the new APIs talk about them as if they will be used by everyone.

Linux at NASDAQ OMX

Posted Oct 22, 2010 6:54 UTC (Fri) by intgr (subscriber, #39733) [Link]

> those programmers aren't going to use Linux specific APIs

Wrong. Those programmers often use proper abstractions that take advantage of OS-specific APIs. For example libevent and java.nio completely abstract away the differences between epoll on Linux, kqueue on *BSD and /dev/poll on Solaris.

And NASDAQ is paying how many kernel hackers?

Posted Oct 21, 2010 18:24 UTC (Thu) by Curan (subscriber, #66186) [Link]

The important information missing IMHO is: how many Kernel hackers are employed by NASDAQ?
It would have a lot of benefits for them; they could get the work needed to improve their use case faster into mainline which would result in less hardware needed and thus less costs (purchase, maintenance, operation (like power)), to name the most prominent one.

And NASDAQ is paying how many kernel hackers?

Posted Oct 21, 2010 19:41 UTC (Thu) by iabervon (subscriber, #722) [Link]

I'd expect them to be paying some other organization that does employ kernel developers, rather than hiring people themselves. I believe we've heard from enterprise distros about work they've been doing for their financial sector customers. I assume that NASDAQ's motive in showing up in person is to let the kernel maintainers know that the scalability work that Red Hat is doing is actually useful in a real-world situation with quantifiable improvements that the customer is willing to talk about; distros sometimes say that something is really beneficial to some customer they aren't allowed to name, whose data they don't have access to, and it's not all that convincing to maintainers who don't know how much it really matters.

And NASDAQ is paying how many kernel hackers?

Posted Oct 21, 2010 19:48 UTC (Thu) by Curan (subscriber, #66186) [Link]

Fair point! I overlooked that.

Thanks,
Cùran

And NASDAQ is paying how many kernel hackers?

Posted Oct 21, 2010 19:53 UTC (Thu) by corbet (editor, #1) [Link]

Actually, an awful lot of these financial institutions do employ their own kernel hackers. Each is trying to cut latencies to the bone, so each chips away on the kernel and they all keep their work to themselves (competitive advantage, after all) and don't talk to each other. I don't know the answer to "how many is NASDAQ employing," but my belief is that it's non-zero.

And NASDAQ is paying how many kernel hackers?

Posted Oct 22, 2010 7:44 UTC (Fri) by conor (guest, #35810) [Link]

& one of the ongoing bugbears for those institutions is ascertaining that their timing metrics are actually as accurate as they can be...

In my experience (in the large institutions) they typically tend have on site SuSE/RedHat engineers to assist with tweaking the kernel (but not normally providing custom kernel builds), rather then hordes of kernel developers employeed unfortunately.

I'm sure there are some institutions that do make more of an investment in them, but they are certainly in the minority.

And NASDAQ is paying how many kernel hackers?

Posted Oct 26, 2010 17:49 UTC (Tue) by Baylink (guest, #755) [Link]

I knew I should have read all the comments before posting mine.

<litella>Nevermind</litella>

Linux at NASDAQ OMX

Posted Oct 22, 2010 9:03 UTC (Fri) by Np237 (subscriber, #69585) [Link]

It’s nice to know that reducing latencies in the kernel can help destroying the World’s economy with High Frequency Trading.

Makes you want to work on other projects.

Linux at NASDAQ OMX

Posted Oct 22, 2010 19:09 UTC (Fri) by i3839 (guest, #31386) [Link]

The proper way to fix that is to disallow such high frequency trading by law worldwide. Start with a one month minimum delay between buying and selling a share to stop the insane speculation that's going on.

Making the Linux kernel slower is not the right fix.

Linux at NASDAQ OMX

Posted Oct 22, 2010 21:28 UTC (Fri) by nix (subscriber, #2304) [Link]

A one month delay is ridiculous: it would essentially ban hedging, which would definitely not improve the health of the financial system one whit.

A one hour, or even one day delay would probably not be so problematic.

Linux at NASDAQ OMX

Posted Oct 23, 2010 16:53 UTC (Sat) by i3839 (guest, #31386) [Link]

It won't ban hedging, that would be still possible. The need for it won't be as much because the whole system would become more stable with less crazy fluctuations and wild speculations. You can still do hedging, everything just goes slower, more sedate, giving people more time to think. The reason for stocks to exist is so that companies can have capital and do things they normally can't do. If you're not interested in investing in a company, then don't buy their stock. But buying and selling again before the company can change or do much is not investing, that's plain speculation or other greedy playing with money.

And yes, something like this will never be introduced because it would be too good for the world and too bad for the people making money by moving it around.

That said, introducing it slowly, starting with one hour, and then slowly moving it up is the only sensible way to introduce it of course.

Linux at NASDAQ OMX

Posted Oct 24, 2010 9:31 UTC (Sun) by i3839 (guest, #31386) [Link]

To clarify, when I say a one month delay I mean for individual shares. If you owned shares in A for at least a month you can sell them any second. Same for buying shares of B. It's only that you're not allowed to sell shares you own for less than a month.

Does this sound more reasonable to you?

Linux at NASDAQ OMX

Posted Oct 24, 2010 15:29 UTC (Sun) by nix (subscriber, #2304) [Link]

Yes... although I can easily think of a way to turn that right around and go back to microsecond-by-microsecond high-frequency trading (you just need a largish position that's 'not old enough to sell yet').

Linux at NASDAQ OMX

Posted Oct 25, 2010 16:25 UTC (Mon) by i3839 (guest, #31386) [Link]

It's probably possible to always bypass things like this one way or the other, so perhaps it just should be encouraged instead of plainly forbidden: Put lower taxes on shares you sell after a long time and tax short lived share more. Though I get the feeling that there might be almost no taxes anyway if such high frequency trading happens so much, so perhaps just increasing the tax on trading shares and lowering the tax on dividend would be a better approach.

Problem of these things is that they only work if it's more or less the same everywhere in the world, or the money moving around will just happen in some other place. Maybe it's a lost cause anyway.

Linux at NASDAQ OMX

Posted Oct 25, 2010 20:04 UTC (Mon) by anselm (subscriber, #2796) [Link]

The German government recently suggested the introduction of an international tax on stock market transactions, which proposal met with resounding silence from just about everywhere else.

Linux at NASDAQ OMX

Posted Oct 26, 2010 20:47 UTC (Tue) by dlang (subscriber, #313) [Link]

there is no tax on a stock transaction, there is tax on the profit from a stock transaction (if you sell the stock for more than you paid for it), but there are exceptions for some cases that were designed to allow a person to move their money from one company to another without loosing money to taxes in the process. The end result is that you pay taxes when you pull your money out of stocks, not when stocks are sold.

Linux at NASDAQ OMX

Posted Oct 28, 2010 3:47 UTC (Thu) by wahern (subscriber, #37304) [Link]

More rules is a horrible idea. More rules means more oversight which means the government and the taxpayers are taking on more of a burden.

IMO the better solution is radical transparency. The account of every trader, broker, and fund should be completely open. We should have bonds, securities, commercial paper and everything else flow through transparent exchanges. We should have an open national database of corporate holdings so you couldn't setup shell companies to hide market transactions.

This way the market could police itself. And when rules are broken they'd be broken by more than one party (as both parties would be responsible for ensuring transparency of the overall transaction), so enforcement of the few remaining rules would be even easier.

Of course, this wouldn't be well received in Europe because of their peculiar views on privacy (gov't invasion okay, private invasions so intolerable as to be immoral), and it wouldn't be well received in America because, well, the corporations wouldn't appreciate not being able to extract rents--and bailouts--from the shoddy "markets" they construct.

Linux at NASDAQ OMX

Posted Oct 28, 2010 13:48 UTC (Thu) by Np237 (subscriber, #69585) [Link]

You are contradicting yourself. How will you enforce transparency, if not by making more rules? If not by adding oversight to control this transparency?

You may disagree on the nature of rules there is to implement but at least you seem to share this view.

Linux at NASDAQ OMX

Posted Oct 28, 2010 19:20 UTC (Thu) by wahern (subscriber, #37304) [Link]

The implication of using transparency as a device is that rather than government officials needing to pour over transactions and to police the minutiae of corporate decision making to detect improper behavior, the market would be able to do it. For example, the market wouldn't have allowed the mortgage bubble to become so inflated if it could have clearly seen the ridiculous positions of the major investment banks and institutional investors viz-a-viz those instruments. As it was nobody knew whether investor A had 5% invested in that particular segment or 95%. Overall you would need fewer government officials and fewer rules.

Financial markets are based on risk, and risk is lack of information. The more information a market has access to, and the less costly it is to leverage that information, the less risk in the system.

Economists would say that the only information a market needs is open pricing signals (i.e. you just need exchanges). But that can't literally be true because you have complex derivative products which manifestly nobody can properly price; and the reason these derivative products are so popular is because of the problem of managing risk. Reduce risk and you reduce the pathologies for pricing at the margins. You also avoid the moral dilemma of too big to fail, because the "too big" institutions couldn't as easily get close to failure mode.

Enforcing transparency is a rule which addresses the illness. Adding ad hoc rules on holding requirements, etc., merely addresses a symptom. Reduce risk and you alleviate--to some extent--all the symptoms.

Linux at NASDAQ OMX

Posted Oct 28, 2010 20:30 UTC (Thu) by Np237 (subscriber, #69585) [Link]

I’m amazed that there are still people to believe in Adam’s Smith invisible hand.

There is no argument that can be opposed to such a religious belief.

Linux at NASDAQ OMX

Posted Oct 29, 2010 17:48 UTC (Fri) by zlynx (subscriber, #2285) [Link]

Do you believe in solving NP-hard problems using techniques such as Monte Carlo and simulated annealing?

"The Market" is just such a technique for optimizing the economy. The global economy (even a country's economy) is a problem impossible to solve with a giant computer or central planning bureaucracy (gag).

Linux at NASDAQ OMX

Posted Oct 29, 2010 18:32 UTC (Fri) by Np237 (subscriber, #69585) [Link]

Monte Carlo and similar techniques do not work for chaotic systems.

And you don’t need such a far-fetched analogy to see that letting “the market” optimize the economy just doesn’t work. Historically, every time “the market” was left with more power and less regulation, it lead to increased poverty and economical crisis.

Linux at NASDAQ OMX

Posted Oct 29, 2010 19:45 UTC (Fri) by zlynx (subscriber, #2285) [Link]

The market isn't chaotic in that sense. Food and water will always be valuable. So will labor reduction (employees, slaves, machines) and entertainment. At a medieval feudal level a planned economy can be run from the top down. That couldn't be done if it was a truly chaotic system.

In the modern world eliminating the market or placing too much control on it will always fail.

The market will resurrect itself in the form of an underground economy where supply and demand (plus legal risk fees) will result in exchanges at the true market value of goods and services.

A regulated market that respects true values will still fail when the regulation cost exceeds the legal risk costs. At that point it becomes much cheaper to avoid the legal markets (pay workers cash under the table, hire illegal aliens, buy drugs from Canada, etc).

Take the old USSR for example or any government that has attempted to enforce a rationing system. Ticket sales and scalpers are another example.

To take this back to the subject, if there was any attempt to impose month long waits on stock sales, investment brokers would go somewhere else, to another country probably, where such ridiculous laws were not in force.

There will always be someone. North Korea or Venezuela could become the finance capital of the world.

Linux at NASDAQ OMX

Posted Oct 29, 2010 20:03 UTC (Fri) by Np237 (subscriber, #69585) [Link]

Economy as a whole is not necessarily a chaotic system. But “the market” with no control (or joke control, as currently) is definitely one.

As for other places becoming the finance capital of the world: sorry to disappoint you, but it has already happened. Bahamas, Cayman, Isle of Man, to a lesser extent Delaware and Singapore… This is where the large transactions are happening. The solution is not to just imitate them and hope for the best when speculation and market manipulation takes away any kind of value from the real world economy. The solution is, on the contrary, to impose strong regulations (with high legal risk costs, as you point out) and to make it costly to make money flee in such places. Several realistic solutions, from Tobin tax to various kinds of oversight, have been proposed in this direction, but have not been implemented because of beliefs in “the market” that confine to religion.

Linux at NASDAQ OMX

Posted Oct 30, 2010 22:43 UTC (Sat) by giraffedata (subscriber, #1954) [Link]

As for other places becoming the finance capital of the world: sorry to disappoint you, but it has already happened. Bahamas, Cayman, Isle of Man, to a lesser extent Delaware and Singapore… This is where the large transactions are happening.

Let's not exaggerate. Far more money changes hands in New York, London, etc. than those places. These are the finance capitals of the world. And it's because laws in New York and London allow exchanges to run Linux and execute a million orders a second. If they didn't, then the transactions would move to those other places.

I don't get the "sorry to disappoint" angle at all, since you're at best confirming zlynx's point: places like North Korea and the Bahamas can have fewer restrictions and be the marketplace of choice.

Linux at NASDAQ OMX

Posted Oct 31, 2010 6:18 UTC (Sun) by Np237 (subscriber, #69585) [Link]

You are completely missing the point.

I don’t care about people playing casino in such places. As long as they don’t do it with our money.

There are ways to prevent people from using money and goods from our countries in such places. Running a competition for which country has the laxest laws to allow the fastest transactions with the least control is not one.

Linux at NASDAQ OMX

Posted Oct 31, 2010 7:04 UTC (Sun) by dlang (subscriber, #313) [Link]

if all you are worried about is your money, then just don't give your money to money managers that do things you don't like.

unless you are forced into a union, they don't force you to give them any money, you can do whatever you want with it.

Linux at NASDAQ OMX

Posted Oct 31, 2010 7:24 UTC (Sun) by bronson (subscriber, #4806) [Link]

Have you forgotten about the great bank bailout of 2008?

Linux at NASDAQ OMX

Posted Oct 31, 2010 7:51 UTC (Sun) by Np237 (subscriber, #69585) [Link]

Do you actually know how the economic system works?

Not everyone of us wants to spend his life alone in the mountains raising goats.

Linux at NASDAQ OMX

Posted Oct 31, 2010 8:13 UTC (Sun) by giraffedata (subscriber, #1954) [Link]

if all you are worried about is your money, then just don't give your money to money managers that do things you don't like.

I don't think anyone is posting worries about his own money, even though the posts might be worded that way out of simplicity. What people are worried about is their wealth, and having a mattress stuffed with money doesn't make you wealthy if the price of everything is high. Certain behaviors of financial institutions make prices high so that no matter what one person does with his own money, he ends up poorer.

And it's rather arbitrary what money we call mine and what money we call yours (laws can easily change that accounting), so it's legitimate to complain about things that waste "our money."

Linux at NASDAQ OMX

Posted Oct 31, 2010 7:59 UTC (Sun) by giraffedata (subscriber, #1954) [Link]

You are completely missing the point.

I don’t care about people playing casino in such places. As long as they don’t do it with our money.

No, I just wasn't addressing that point. I was addressing the point from your immediate previous post that the Cayman islands, etc. are today the finance capitals of the world, and that that revelation should disappoint zlynx (who said places like North Korea could become finance capitals if laws were to restrict trade in today's finance capitals).

If those weren't points you intended to make, you shouldn't have said it.

Linux at NASDAQ OMX

Posted Oct 31, 2010 13:07 UTC (Sun) by nix (subscriber, #2304) [Link]

I wish people would pay attention to the history of the Soviet Union instead of just using it as a whipping boy. What killed their dream of near-optimal central planning from the start was two things: firstly, the available computer power and communications bandwidth was nowhere near sufficient to optimize the whole economy, or to react to changing external conditions fast enough; secondly, they relied on people reporting what was happening on the ground, and most of those people had an incentive to *lie*. (Later on, a third thing killed them: that the existing productive systems had grown into empires and the people running those empires fought to stop them being torn down: also, by then the lies were so pervasive that not many people thought to tear them down. Capitalism's "creative destruction" may not be all it's given credit for but it's a hell of lot better than what they had.)

But the idea of optimizing an entire economy is not hopeless. You just can't do it with humans doing the data collection and reporting :/ in this it was very similar to much else about communism: it would work great if we were all selfless robots.

The nice thing about price signals is not that they are particularly brilliant or insightful. It is that they are relatively hard to manipulate, so individual liars can't bugger up your economy.

Linux at NASDAQ OMX

Posted Oct 23, 2010 20:27 UTC (Sat) by kleptog (subscriber, #1183) [Link]

Seems completely pointless to me. At that would happen is that a new derivative would be created where someone promises to sell shares for at a particular price now and you will sell the same amount to them in a hours time. Shares are fungible. If you have 1000 shares in X and you buy 100 now and sell 200 the next second, have you sold the same shares or different ones?

The problem is not the trading itself, it's people taking risks they shouldn't. This is a social problem that will not yield readily to a technological solution.

Linux at NASDAQ OMX

Posted Oct 24, 2010 9:29 UTC (Sun) by cmccabe (guest, #60281) [Link]

> Seems completely pointless to me. At that would happen is that a new
> derivative would be created where someone promises to sell shares for at a
> particular price now and you will sell the same amount to them in a hours
> time. Shares are fungible. If you have 1000 shares in X and you buy 100
> now and sell 200 the next second, have you sold the same shares or
> different ones?

Well, you could imagine a system where trades are only possible every 60 minutes. Presumably everyone would submit his order, and they would all get executed at once. I guess people could always choose to trade their shares on a dark pool or something, though.

Apparently the SEC has publicly declared that high-frequency traders had no part in the 2010 Flash Crash.

> The problem is not the trading itself, it's people taking risks they
> shouldn't. This is a social problem that will not yield readily to a
> technological solution.

Well, there's really a few different problems that caused the credit crunch. The first, and probably most serious, is that the U.S. government allows asset bubbles to form and explode without doing anything. Whenever China's stock market starts overheating, they introduce new regulations, start limiting price movements, etc until the problem is under control. Similarly, with property prices, China takes steps to keep them from overheating. For some reason, it's accepted wisdom that the U.S. government has to intervene when things are bad, but not keep a lid on things when the economy starts overheating. It's weirdly asymmetrical.

So we had kind of a "finance bubble" in general. The whole business of moving around money became one of the hottest sectors of the economy, much like computer technology in the late 90s. Exotic derivatives were developed-- what Warren Buffet called "financial weapons of mass destruction." In a lot of cases, human judgement was replaced by computer heuristics and algorithms developed by "quants."

The results, I guess you can see. But it wasn't greed or excessive risk-taking that caused the problem. Just a regulatory and conceptual failure. Remember, it's possible for everyone to act rationally, and yet to end up with a deeply irrational outcome.

Linux at NASDAQ OMX

Posted Oct 24, 2010 21:10 UTC (Sun) by butlerm (guest, #13312) [Link]

The real conceptual failure is allowing banks to lend out "deposits" in the first place. If banks were required to keep 100% reserves on deposits (as the term implies), the banking system wouldn't always be one incident away from total collapse.

Linux at NASDAQ OMX

Posted Oct 24, 2010 23:22 UTC (Sun) by anselm (subscriber, #2796) [Link]

The problem with this is that deposit customers expect the bank to pay them interest, which the bank earns by lending other folks the deposited money at a higher interest rate than what it pays to the money's actual owner.

If banks were required to hang on to all deposits in full, there would not only be no point in banking your money if the bank wasn't able to pay you interest. It would also be very difficult for you to get a loan because the bank would have to borrow money elsewhere to give to you – and it can't borrow from other banks because they have no money to lend, either, because they need to hang on to their deposits, too. Which kind of throws a spanner in the works of an economy.

The banks could of course try to borrow from the national central bank, which would theoretically be in the unique position of being able to deal with the demand by printing »fresh« money, but that would increase the amount of money in circulation and lead to inflation, which the national central bank would not like (and which would devalue all the no-interest deposits in the consumer banks), so you're stuck.

Linux at NASDAQ OMX

Posted Oct 25, 2010 19:46 UTC (Mon) by butlerm (guest, #13312) [Link]

The problem with this is that deposit customers expect the bank to pay them interest

It is metaphysically impossible to earn interest (in real terms) on a "deposit". Loans earn interest, deposits don't. The whole point of a deposit is to sit there, until you need to withdraw. The problem is that banks started lending out deposits (illegally I might add) hundreds of years ago, thereby changing what they still call a "deposit" to a "loan" (to the bank) while still pretending that these "deposits" were risk free transactions.

Banks do not need to be able to lend out deposits in order to be able to lend. They can raise money in the bond market and lend that out instead. They do that now. Lending out deposits just places the rest of us on the hook for preternaturally self destructive behavior.

Linux at NASDAQ OMX

Posted Oct 26, 2010 0:40 UTC (Tue) by cmccabe (guest, #60281) [Link]

>> It is metaphysically impossible to earn interest (in real terms) on a deposit".

Hey, there's nothing metaphysical about it. The interest Bank of America offers on its savings accounts these days doesn't even beat inflation.

And people wonder why Americans never save money...

Linux at NASDAQ OMX

Posted Oct 26, 2010 2:10 UTC (Tue) by marcH (subscriber, #57642) [Link]

> And people wonder why Americans never save money...

I think "people" wonder why Americans have huge debts, both as individuals and as a country.

Linux at NASDAQ OMX

Posted Oct 26, 2010 7:37 UTC (Tue) by butlerm (guest, #13312) [Link]

It has nothing to do with the Bank of America. The Federal Reserve is in the business of manipulating interest rates to anything they think is best for the economy. Right now that means essentially zero interest.

Philosophically speaking, however, I don't see why anyone should expect to earn interest on a risk free transaction. Any interest you earn (in real terms) on an FDIC guaranteed savings account comes by courtesy of a federal guarantee, a guarantee that will have to be made good by the taxpayers every generation or so. Banks, as we know them today, are fundamentally unstable institutions. Earn interest today, pay for it in increased taxes (or inflation) tomorrow.

Linux at NASDAQ OMX

Posted Oct 26, 2010 10:45 UTC (Tue) by tialaramex (subscriber, #21167) [Link]

"Philosophically speaking, however, I don't see why anyone should expect to earn interest on a risk free transaction."

I don't think there's any philosophy involved here, just economics.

If you assume (as most of the world does) that progress is not an illusion, then making idle resources active in pursuit of progress is a gain. Remember that to make these transactions "risk free" to the saver we need only in fact make them less risky in total than keeping your money under the mattress. The tricky bit (which is handled by things like FDIC) is smearing the uneven risks so that ordinary citizens don't get a nasty surprise.

"Earn interest today, pay for it in increased taxes (or inflation) tomorrow."

You assume a zero sum where there is no evidence for one. The permanent building societies make a mockery of this claim. Since the early 19th century they've been using Peter's money to buy Paul a house, then charging Paul interest and paying it to Peter, they've put vast numbers of people into homes, given back vast sums in interest, and they're cheap to run. It's just organised sharing, the kind of thing we try to teach kids in kindergarten.

However mutual societies don't make people obscenely rich. And so in the 1980s and 1990s this created an enthusiasm for demutualising the societies (people trying to capitalise on this were called "carpetbaggers") and "unlocking" their assets by converting them into banks. The surviving building societies (which fought off carpet baggers with a variety of tactics) are still going today, their worst exposure to the financial crisis being an increase in defaulters which is easily covered by their mandatory long-term holdings. In contrast many of those which demutualised are now in ruins, being rescued by the government, or by government deals struck with larger foreign banks. The carpetbaggers (ordinary people, albeit greedy ones) keep their loot and the tax payer picks up the bill.

Linux at NASDAQ OMX

Posted Oct 29, 2010 6:53 UTC (Fri) by butlerm (guest, #13312) [Link]

"You assume a zero sum where there is no evidence for one."

I am referring to systems comparable to the contemporary U.S. banking system: fractional reserve, fiat currency, government guarantee of "deposits". The first property of any such banking system is the real interest rate earned by savers (prior to government overhead) is the nominal rate minus the inflation rate. The real rate is *much* smaller than the nominal rate.

During the 1970s it was sharply negative, in other words savers lost money in real terms every year, because inflation was so high. Most of the time the real interest on government guaranteed savings accounts is one percentage point at best. Whether the the $200 billion we are churning out to sustain Fannie Mae and Freddie Mac is enough to reduce government guaranteed savings deposits to a true zero sum game is hard to say. There is no question that they are _close_ to a zero sum game.

There is a perfectly good way to accomplish what you are suggesting without fractional reserve _banking_ or government guaranteed deposits. Have individuals invest in diversified bond funds that purchase bonds issued by building societies, credit unions, and the like. People still get loans to build homes, investors earn more in real terms, and the rates automatically and properly float with the risk the bond issuers undertake - instead of as now where you earn half a percent in real terms whether your bank is about to go under or not, which is a very high price to pay for a taxpayer guarantee.

Poorly managed building societies might fail (affecting the return on bond funds), but the commercial banking system as a whole will be immune from systemic failure, because true _deposits_ will no longer be lent out, but rather only the proceeds of what are in effect loans to independent lending institutions, with the risk widely distributed, instead of concentrated with those so unfortunate to make a direct loan to a bank and/or the taxpayers that guarantee whatever crazy scheme the banks will cook up next.

Linux at NASDAQ OMX

Posted Oct 30, 2010 16:12 UTC (Sat) by kleptog (subscriber, #1183) [Link]

I suppose whether it's a zero-sum game depends on what you're measuring. The economy now is vastly larger than it was 30 years ago. On average everyone is wealthier.

One important thing to note is that the total amount of money in the system is not the important factor (though it is the easiest to control). What is important is the *flow* of money, represented by the velocity of money (how money times is the average dollar spent in a year). Actually, it's the goods and services that the money is exchanged for that is the real economy, but the flow of money is the easiest way to measure it.

(Yes I'm glossing over the various types of money, that not really relavent here).

Money that isn't moving is dead. Money under your bed or in a bank that's not allowed to use it is essentially non-existent for the economy.

So if everyone is spending money like mad, you don't need very much real money in the system to support it. Conversely, if everyone stops spending all of a sudden, you need to make more of it to avoid your economy grinding to a halt.

(BTW, in NL the savings are guaranteed by the other banks in the country. In return they end up the queue of creditors in place of the original customers. Sounds like a nice idea, until the amount guaranteed exceeds a significant amount of the total savings in the country, in which case you're still fucked.)

Incidentely, this explains the problem with your solution. Because you are no longer allowed to use deposits to finance other things, the governments would need to print a *lot* more money to keep the economy the same size. I'd guess about 20 times what is currently in circulation (you can't use substitutes like gold for savings, because its value isn't constant). At any particular moment a significant chunk of the total cash in circulation would be stuck in warehouses, rusting.

Suppose someone sells a house and puts that money in a bank account. This money is for all intents and purposes removed from the economy. To keep everything going the government would need to print that amount again to keep the total up, and also destroy it again when they do decide to spend it. That requires a lot more government intervention (and crystal ball gazing) that what we have now.

Linux at NASDAQ OMX

Posted Oct 31, 2010 13:23 UTC (Sun) by nix (subscriber, #2304) [Link]

On average everyone is wealthier.
In the last 30 years? In the US and UK this is only true because the rich have got so much richer: the rest of us have barely gained a thing.

Linux at NASDAQ OMX

Posted Nov 4, 2010 10:39 UTC (Thu) by linuxrocks123 (subscriber, #34648) [Link]

I don't understand this view, though I've heard it many times. Do you really believe that the average person in the U.S. is no better off than 30 years ago? Were personal computers, safer automobiles, cell phones, the Internet, and portable music players really so minor in improving people's lives that their benefit was negated entirely? I see a lot of ways in which the lower and middle classes of developed countries are better off today and few if any ways they are worse. Can you elaborate on why your perspective is so radically different from mine?

---linuxrocks123

poverty gap

Posted Nov 6, 2010 16:57 UTC (Sat) by tialaramex (subscriber, #21167) [Link]

There is a growing dollar gap between rich and poor. In many industrialised countries the rich have proportionally more money than they did thirty years ago. This worries some people a lot. Even if they happen to be rich, it could be worrying because they sense injustice, or because they fear violent insurrection.

However you and I are interested in the real economy, which is not concerned with money, but with the goods and services which we obtain in exchange for our labour. So we see a different picture. Everybody has better goods and services than thirty years ago, and the richest can afford only very slightly better toys and nicer services than the poor, despite their bulging wallets. Ordinary working people today may easily obtain things that were beyond the reach of the richest people alive a few decades go.

Yet a third way to look at it is happiness. People seem to be not much happier than previous decades, and wealth or poverty seem to have little impact so long as a person is not so very poor that they lack food and shelter and fear for their life.

Linux at NASDAQ OMX

Posted Oct 24, 2010 20:09 UTC (Sun) by marcH (subscriber, #57642) [Link]

> Seems completely pointless to me.

What's really pointless is the Global Gambling Markets, acting on totally surreal time and money scales. Did you know that traders "working" there have poker classes for training?

> Shares are fungible.

Just make them not. Attach a timer to each single share and problem solved.

Basically every single activity in this world is regulated and taxed, except finance. Not for any sensible reason but just because money and power are too well connected.

Linux at NASDAQ OMX

Posted Oct 24, 2010 23:34 UTC (Sun) by nix (subscriber, #2304) [Link]

Uh, finance is most definitely regulated and taxed, and if you don't do what the regulators say they can do quite considerable violence to you.

It's just that the regulators were asleep for a while and the clever-clogs financiers figured out ways around the regulations. You'll note they didn't openly *flout* what regulations there were unless they thought nobody was watching: it's too dangerous even for the likes of Goldman Sachs to openly break the law. Easier to get it changed.

Linux at NASDAQ OMX

Posted Oct 26, 2010 17:55 UTC (Tue) by Baylink (guest, #755) [Link]

> It's just that the regulators

repealed Glass-Stegall, which kept this from happening for, oh, I dunno, 70 or 80 years...

there, FTFY.

Linux at NASDAQ OMX

Posted Oct 29, 2010 15:43 UTC (Fri) by nix (subscriber, #2304) [Link]

I was thinking from a UK perspective, where we didn't actually loosen the regulations: we just had a famously 'light-touch' regulator (the FSA) which turned out in the end to be very much too light with its touch, allowing major banks to finance themselves entirely on the wholesale market and take on exposures to massive amounts of US subprime risk until we had the first bank run for generations and an emergency government takeover.

Linux at NASDAQ OMX

Posted Oct 25, 2010 7:37 UTC (Mon) by kleptog (subscriber, #1183) [Link]

People seem to be under the impression that stock markets are special and strange. It's just a variation of contract law, where people buy and sell things from each other under agreed rules. All stock markets do is have standardised rules so the whole lot can be automated.

Once you go off market, you can invent anything that's not standard. Many of the transactions involving the new derivatives that went wrong had nothing to do with stock markets and everything to do with contracts with new and unusual terms. They weren't transacted at high speeds, they involved months of negotiation between the involved parties (with lawyers). There are rules, but as with contract law in general, if the parties agree then its binding.

The irony is that people *demand* growth in their pension funds, bank accounts, low interest rates on home loans, etc. Yet they don't want to run any risk. The only zero risk option is keeping your money under your bed. It will deflate over time, but it's a *guaranteed* deflation.

The question is not, do people take risks, the question is how much risk do you accept. This is the real world, there is not a lot of history to learn from and sometimes we get it wrong. We pick ourselves up, and try better next time. Until the next technological breakthrough throws a spanner in the works.

Linux at NASDAQ OMX

Posted Oct 25, 2010 11:20 UTC (Mon) by marcH (subscriber, #57642) [Link]

> People seem to be under the impression that stock markets are special and strange.

So far no one has ever seen "high frequency trading" happening anywhere else. Among others, taxes would nip that in the bud.

> The irony is that people *demand* growth in their pension funds, bank accounts, low interest rates on home loans, etc. Yet they don't want to run any risk.

Finance complexities are beyond the reach of "normal people". Especially when products are complex on purpose for the sole purpose of hiding risk (or else). The best people can do is trust their bank/insurance/pension fund based on tie colour. As long as the industry explains in simple and *honest* profit/risk terms it can work fine. This did not happen because this industry was too busy playing poker and trying to cheat each other with products even themselves could not understand any more. When the whole industry goes mad and greedy and *lying* about incredible profits and near-zero risk, and when corrupted regulators and politicians go "asleep", what could people seriously have done? Nothing really. The only thing they can do now is hanging their politicians to try to teach them a lesson.

Linux at NASDAQ OMX

Posted Oct 26, 2010 17:56 UTC (Tue) by Baylink (guest, #755) [Link]

> People seem to be under the impression that stock markets are special and strange. It's just a variation of contract law, where people buy and sell things from each other under agreed rules.

What other commercial transaction risks your job? (As an independent third party, I mean)

Linux at NASDAQ OMX

Posted Oct 26, 2010 3:17 UTC (Tue) by eduard.munteanu (guest, #66641) [Link]

That insane speculation is pretty much the reason for market liquidity and meaningful prices. Putting barriers to it doesn't improve volatility, have a look at low trading volume markets and see the huge ask-bid gaps and sudden variations.

Linux at NASDAQ OMX

Posted Oct 26, 2010 8:11 UTC (Tue) by marcH (subscriber, #57642) [Link]

So what?

You are not the first one basically saying: "you cannot slow down transactions, because it would... slow down transactions". Well, that is the point.

Ses also: http://en.wikipedia.org/wiki/Tobin_tax

Linux at NASDAQ OMX

Posted Oct 28, 2010 22:13 UTC (Thu) by martinfick (subscriber, #4455) [Link]

Slowing down transactions slows down value resolution. The better value can be resolved, the more efficient a global economy can run. The slower value can be resolved, the more "wrong" that information will be. The more people make resource allocation decisions based on bad info, the more bad decisions are likely. This is simple economics. Like it or not, ethics aside, you cannot sacrifice resolution without sacrificing efficiency.

Linux at NASDAQ OMX

Posted Oct 29, 2010 7:18 UTC (Fri) by Np237 (subscriber, #69585) [Link]

This is purely a dogma. It has never been proved by anything else by self-called “economists” who rely on the assumption that their economical model is right.

Linux at NASDAQ OMX

Posted Oct 29, 2010 7:18 UTC (Fri) by Np237 (subscriber, #69585) [Link]

I meant “anything else but”

Linux at NASDAQ OMX

Posted Oct 29, 2010 16:52 UTC (Fri) by martinfick (subscriber, #4455) [Link]

I agree, ignore the economists, most of them are politicians.

Nevertheless, which part of my statement/implications are you calling dogma? ...and please explain.

1) Economic value changes
2) Out of date information about a changing value could be bad information
3) Decisions based on inaccurate information have a greater chance of being bad decisions than those based on accurate information
4) Economic value information influences resource allocation
5) Less accurate information about economic value will lead to less accurate resource allocation
6) Less accurate resource allocation prevents more efficient resource allocation

Linux at NASDAQ OMX

Posted Oct 29, 2010 18:30 UTC (Fri) by Np237 (subscriber, #69585) [Link]

I agree that if information is good, people will make better decisions, from their point of view.

The dogma is that if each actor makes his own, well-informed decisions, according to his own interests, the economy as a whole will work in an optimal way.

Such theories have been proved wrong again and again, whether in the real-world economy, or in similar mathematical and physical phenomenons. As a chaotic system, the economy as a whole cannot be predicted nor optimized.

Linux at NASDAQ OMX

Posted Oct 29, 2010 19:09 UTC (Fri) by martinfick (subscriber, #4455) [Link]

> The dogma is that if each actor makes his own, well-informed decisions, according to his own interests, the economy as a whole will work in an optimal way.

I could see how it would seem like I was saying that, sorry. To rephrase your words closer to something closer to what I did mean:

The economy as a whole will likely work in a LESS optimal way if each actor makes his own, according to his own interests (how can they do otherwise?) LESS informed decisions, than if those decisions were better informed.

In other words, I am not pinpointing the road to optimality, but I am pointing out a road which can clearly be deduced to point away from it.

Linux at NASDAQ OMX

Posted Oct 29, 2010 19:23 UTC (Fri) by Np237 (subscriber, #69585) [Link]

That, I can agree with. More transparency, from all companies and public institutions, can only help.

Linux at NASDAQ OMX

Posted Oct 29, 2010 15:45 UTC (Fri) by nix (subscriber, #2304) [Link]

But we also see increased volatility when high-frequency traders get involved.

It is reasonable to presume that there is a low-volatility sweet spot where trading is neither too fast nor too slow/thin, but as far as I know nobody has even proposed looking for one, let alone figuring out how to approximate the location of that spot or figuring out how it varies (and even along which dimensions it exhibits variance).

Linux at NASDAQ OMX

Posted Oct 29, 2010 17:25 UTC (Fri) by martinfick (subscriber, #4455) [Link]

The volatility is there, whether you let people trade or not. In other words, if my cell phone becomes obsolete overnight or not is not relevant to whether I am allowed to sell it or not. You can mask volatility by not allowing me to trade, but you cannot eliminate (or even change) it. Pretending my cellphone is not obsolete, by preventing me from selling it as soon as I realize it is obsolete, may make the market look less volatile, but it is a coverup, a lie, no one benefits. If I can't sell it, I am robbed of an opportunity, so is my potential buyer, who else could potentially benefit from our mutual loss? Lies may make some people feel better, they do not make me feel better. These types of lies are highly likely to make the economy less efficient, and likely to waste a great deal of resources.

Linux at NASDAQ OMX

Posted Oct 30, 2010 17:40 UTC (Sat) by jpnp (subscriber, #63341) [Link]

Now I can well imagine that if we regulated cell phone transactions so that, say, you could only buy and sell your phone on one day of the month, then you and your potential buyers would lose out and fail to allocate your resources efficiently.

I'm not convinced that this can be extrapolated to trading times less than a millisecond (way faster than the human mind).

Linux at NASDAQ OMX

Posted Oct 30, 2010 18:22 UTC (Sat) by martinfick (subscriber, #4455) [Link]

What does the human mind have to do with it? If a certain price makes purchasing a product suddenly cost effective, then why should I (or my computer) be prevented from buying that product just because that price only stayed cost effective for milliseconds? This is not a hard scenario to imagine. Perhaps phones cost $70 from the warehouse. If some warehouse has a glut suddenly (they need to move them out to make room for something else), and they offer them for $65. The first (fastest) buyer/retailer gets the lot. Surely it would be unjust to limit this sale.

But take this further, as soon as that transaction has been performed, another one could potentially become valuable! Many transactions only become valuable because of several other transactions. In other words, the reason the warehouse may have wanted to make the sale as quickly as possible at $65 (even though perhaps they could get $67 if they waited longer) is because they were faced with a similar deal on the other end; that is why they needed to empty their warehouse of the phones. If they could not sell the phones, they might not be able to purchase the other product in time at the right price. They would lose out, and so would their supplier and buyer, everyone in the chain looses! As you can perhaps extrapolate, this becomes a large global chain of best deals. The faster they are made, the more efficient the economy can operate, any delay is a potential waste of resources.

If there is one thing that open source should have made clear is that even minute barriers become problematic in a global collaborative environment. Those open source projects which understand this and work hard to minimize barriers are those which are most likely to succeed (because they are likely more efficient). Imagine if someone unrelated to your project, thought that it were OK (and they had the power) to limit the rate of contribution to your favorite project, would it not be unjust? Time and time again this is questioned, wikipedia vs. nupedia..., and time and time again, those with the lowest barriers displace those with higher barriers. This is not just theory, this is science, look at the empirical data, not some economist's or politician's analysis.

Play "Settlers of Catan" a bit with a competent crowd and you will quickly realize how valuable it is to respond quickly to offers and also how valuable it is to make them quickly! :)

Linux at NASDAQ OMX

Posted Oct 31, 2010 13:03 UTC (Sun) by nix (subscriber, #2304) [Link]

You are assuming that all trades high-frequency-trading algorithms make are useful. Have you *seen* what those things do? Millions of tiny trades of the same stocks backwards and forwards to capitalize on tiny shifts in price seem reasonable until you realise that these things are all running secret algorithms and nobody's figured out their ensemble behaviour. And their ensemble behaviour fairly often (from our perspective: rarely, from their millions-of-trades-an-hour perspectie) seems to include huge bursts of buying or selling that no sane human would ever have asked the bots to do.

i.e. we have reached a situation where bugs in a massively-distributed system written by thousands of hands without any coordination whatsoever can immediately destroy chunks of the financial system. Kludge tripwires have been welded into the exchanges to try to stop it, but all these can possibly ever do is ameliorate the problem.

Linux at NASDAQ OMX

Posted Oct 31, 2010 18:37 UTC (Sun) by martinfick (subscriber, #4455) [Link]

> You are assuming that all trades high-frequency-trading algorithms make are useful.

No, I don't see any assumption of the sorts in my post(s). I am pointing out that by banning fast trades, you will be placing clear barriers that will prevent the economy from being more efficient. Could plenty of other things make the economy less efficient? ...Sure. But you simply cannot limit trades and not expect to adversely impact things.

> i.e. we have reached a situation where bugs in a massively-distributed system written by thousands of hands without any coordination whatsoever can immediately destroy chunks of the financial system.

Hmm, I suspect that prices are the great invisible coordinator that you seem to be forgetting about.

> Kludge tripwires have been welded into the exchanges to try to stop it, but all these can possibly ever do is ameliorate the problem.

If those exchanges want to provide "fixes" to their customers, great (as long as other exchanges are free to do otherwise). But, those exchanges which do a better job at preventing barriers to trading in the long run are those which people are more likely to use in the long term.

Linux at NASDAQ OMX

Posted Oct 31, 2010 19:18 UTC (Sun) by nix (subscriber, #2304) [Link]

Uh, the 'great invisible coordinator' of prices... doesn't work. There have now been multiple examples of HFT systems causing extremely radical swings in both trading volume and price, because unanticipated interactions between different systems lead to both choosing to buy or sell or whatever until they've driven the price through the roof or floor. (On top of that, to an HFT system, virtually every market is illiquid because they dominate the trading volume of every market they arrive in sooner or later, and illquid markets are famously hard to price accurately.)

But perhaps you'd rather prefer to believe that prices are mystically perfect, when they plainly are not. Not remotely.

And this is really seriously off-topic, so this will be my last post in this thread.

Linux at NASDAQ OMX

Posted Nov 1, 2010 16:39 UTC (Mon) by marcH (subscriber, #57642) [Link]

> The volatility is there, whether you let people trade or not. In other words, if my cell phone becomes obsolete overnight or not is not relevant to whether I am allowed to sell it or not.

Is this a disguised car analogy? It does not look any better than one.

Linux at NASDAQ OMX

Posted Oct 28, 2010 21:09 UTC (Thu) by jhhaller (subscriber, #56103) [Link]

The high volume transactions are not always exercised, they can be canceled before be executed. One could see this used as an offensive weapon, by sending in many small orders which won't be executed, but swamp your competitor's ability to see what are the real market prices. There is a small tax on executed trades (charged to the seller), so there is some financial incentive to minimize the number of successful trades, particularly for small gains. Of course, with NASDAQ being in the middle of this, they need to be even faster than the broker/dealers making and/or canceling trades.

½OT: LSE switching over to Linux

Posted Oct 25, 2010 14:56 UTC (Mon) by Curan (subscriber, #66186) [Link]

Just read, that the London Stock Exchange is now also running on Linux (after they were featured prominently in MS ads in the past, at least over here) and trumpeting "world record trading speed" (126 microseconds/transaction).

½OT: LSE switching over to Linux

Posted Oct 30, 2010 1:06 UTC (Sat) by mab (subscriber, #314) [Link]

½OT: LSE switching over to Linux

Posted Nov 1, 2010 8:58 UTC (Mon) by buchanmilne (guest, #42315) [Link]

It seems they switch over today:

The news comes ahead a major Linux-based switchover in twelve days, during which the open source system will replace Microsoft .Net technology on the group’s main stock exchange.

[...]

Millennium Exchange is set to go live on the LSE’s main exchange in 12 days’ time, on 1 November.

Linux at NASDAQ OMX

Posted Oct 26, 2010 17:47 UTC (Tue) by Baylink (guest, #755) [Link]

> What NASDAQ OMX would really like to see in Linux now is good socket-based AIO.

So, how many Linux developers are paid employees of NASDAQ?

I'm not trying to be snotty there, either; that's the primary attraaction for lots of big users to Linux, and I don't see that this piece says -- that would seem to be about the 3rd or 4th question I would have expected...


Copyright © 2010, Eklektix, Inc.
This article may be redistributed under the terms of the Creative Commons CC BY-SA 4.0 license
Comments and public postings are copyrighted by their creators.
Linux is a registered trademark of Linus Torvalds