More Warnings About High-Frequency Trading 500
bfwebster writes "From The Big Picture (a great finance/econ blog) comes a link to this New York Times article on some of the risks and problems of high-frequency trading on financial markets and a couple of 'gadflies' who are pushing hard to get some changes and reforms in how Wall Street handles HFT. Key question: when is fast trading too fast?"
Re:it's too fast (Score:5, Interesting)
Wish I had mod points for the parent here.
The key point HFT people keep harping on is increased liquidity. The issue is that at some point, a point we reached long ago, that increased liquidity became meaningless to investors and the HFT tax became a burden. Even institutional investors pay this tax. Only other HFT traders care now.
They are skimming money off the top, adding ZERO value to the market, and pissing off just about everyone.
But they have the money, so they can prevent regulations from limiting this... (plus regulaiton are always bad, right?)
Zero Sum, so where does the money come from? (Score:4, Interesting)
At such short timescales, trading is a provably zero-sum game. So where do all the fantastic profits that HFT operations claim come from? Everyone else. If you invest in a stock, during that process, an HFT algorithm (or ten) attempt to manipulate the market to cost you a fraction more, sweating the coins [wikipedia.org] that you might receive. (The rest of the time, the HFT algorithms end up fighting each other, but apart from driving the market unstable, its only the HFT operators who win/lose amongst themselves, the HFT industry gains nothing).
Yet they don't actually provide the much vaunted "liquidity": if they did, they couldn't extract the revenue by making the liquidity dissipate when its actually needed: if the HFT bots added liquidity, Knight Capital wouldn't have taken a huge loss, as they could have sold the stock they bought back to the market rather than having to lose $400M! selling the shares to Goldman Sachs.
It really is time for a microscopic but non-zero Tobin tax [wikipedia.org] on stock transactions: $.00001 per buy or sell request issued to the market. That should stop the bots from spamming the market with bogus requests, and level the playing field for everyone else.
Re:Speed doesn't matter (Score:5, Interesting)
Except that if you have sell triggers set based on normal movements of a stock and it happens to get caught up in a flash crash, it could easily execute at the low price then bounce and you'd never have time to react.
goals and chaos theory (Score:5, Interesting)
Here's what most magazines and newspapers discussing the topic are missing:
A definition of what we want the stock market to be like.
Everyone is focussing on what they don't want. But that's not how you build a resilient system. Basically, that's using default-allow for your firewall. You'll be spending the rest of your life adding rules of what you don't want.
Once you switch around your mind, the questions become a lot easier. Decide what you want the stock exchange to be, and you get your answers almost for free.
If you want the stock exchange to be a place where companies can meet investors and get capital raised, then everything that doesn't serve that purpose directly or indirectly is out. You define how the process should work and allow only that, done. Everyone who wants to play games will have to do it within the parameters you have defined.
The whole problem here is that too many people still believe the old nonsense about the invisible hand. Yes, to some extent you can build a sandbox and people will come and build their sand castles. You can provide a market place and have the participants sort out how everything works.
But you will get scammers, fraudsters, thieves, HFTs and all the other scum as well. If your sandbox is an MMO, you will get gold farmers and scammers and spammers. If your sandbox is a stock exchange, you will get HFTs and stock fraud and insider trading.
Letting chaotic self-organization create the rules of the game through emergence is an interesting experiment, one that I enjoy quite a bit when it comes to games or small settings (book a weekend with friends in a summer cottage and something will happen, no need to set up a schedule beforehand).
Allowing corrupt idiot politicians to base the world economy on chaos theory was one of the dumbest ideas we as a species ever had. Read some catastrophy theory [gold-eagle.com] first (at least check out the graphic if the article is tl;dr). There's a reason we call it chaotic systems, you know?
Content free (Score:5, Interesting)
The NYT article sucked. Can't wait for print journalism to die.
1) Here's some authorities who are authorities because we say so, who are fighting like little middle school drama queens
2) Everyone loves a good "rich people suck and they're corrupt
3) Rabble rousing tired old cliche of "kids have no idea what they're doing vs old people are obsolete"
If you want the real story try Stucchio's blog series beginning at:
http://www.chrisstucchio.com/blog/2012/hft_apology.html [chrisstucchio.com]
My pitiful TLDR summary of Stucchio's work combined with some other observations
1) Idiotic SEC rule 612 quantizes share price into small enough increments that you can raise the capital to HFT and increments that are large enough that you can make some serious dough doing HFT. HFT is not caused by Bolshevists in your bathroom or too many atheists or gobblin infestations, its caused by a stupid SEC rule that was created in a snap of the fingers and can go away just as quick.
2) Ask ANY EE or "real" telecom guy HFT is a real world application of dithering to improve resolution. Oh the "real" price of a share is 100.003 but we're only allowed to report price to a resolution such that its either of two quantum states 100.00 or 100.01. Simple system solution? Dump out 10 orders, 3 at 100.01 and 7 at 100.00 to meet some idiotic SEC rule. Now "everyone" knows the real free market price is signalled at 100.003 even if the unfree market is only allowed to trade at one cent increments.
3) Most of the whiners on both sides have a dog in the fight, if they are in the business and can't HFT for whatever reason they hate HFT and if they can they love it. Or they're just doing witchcraft style persecution where no logical mechanism is necessary... my sheep died therefore we should hang some old woman is no different than my dotcom bombed therefore we should punish a successful HFT trader "just because".
4) The ratio of HFT trading vs retail trading is absolutely exploding not entirely because HFT is growing (although it is...) but because retail trading is absolutely dying. Retail is dying for two reasons: Headed into next down leg of the second great depression (bubbles usually melt up in price and down in volume) and demographic stuff like looking at labor force participation graphs tens of millions of working age americans are no longer working, therefore there is no need to invest their 401K and IRA money from non-existent jobs. So yes HFT is growing but don't make the mistake of thinking the ratio of HFT to retail is in any way relevant, because retail is terminally ill and almost dead.
5) Some elderly/lazy people get all confused about frontrunning (which is illegal) because both frontrunners and HFTers are nuts crazy about low latency and fast execution. Therefore they obviously must be the same, at least to an idiot. This is about as intelligent as the anti-vaxers ... "Getting hit in the head by a 2x4 causes stupidity... high levels of lead in the blood cause stupidity... therefore getting hit in the head by a 2x4 results in high blood levels of lead". Morons.
6) Some lazy people enter market orders instead of limit orders. Probably not a wise idea any time in the last 150 or so years, although it is an even worse idea when retail volume is melting down and HFT volume is melting up. You could ban market orders for retail investors, but stupid people are always going to find a way to lose money, so I'm not sure there's any point to it. There's an infinite pool of ways for morons to lose money so removing one isn't going to really change the outcome. Its not the end of the world in that any average long term retail investor trade has just about zero odds of being stuck in a "flash-anything" related HFT event.
Re:Trading's Too Fast When It Ceases to Mean Anyth (Score:4, Interesting)
Re:High Frequency trading (Score:5, Interesting)
This is why I favor a 0.1-0.5% fee. If we really want, we could also assess the fee as an inverse function of how long you've held it: the longer you've held the stock, the lower the fee, and it gets exponentially smaller with time until it is insignificant.
But if you want to trade something you bought 5 seconds ago? Enjoy your 90% fee.
Re:Trading's Too Fast When It Ceases to Mean Anyth (Score:5, Interesting)
The rate at which these decisions are being made indicates that it is not going through a human mind.
Why's that a problem? Index trackers don't involve human minds either.
I'm fine with HFT. My only conditions are:
1) No rollbacks for HFT trades[1]. You screw up you eat the loss.
2) If bailouts are needed for whatever reason (your company loses billions of other people's money), the traders involved (if any) and the bosses go to jail for 20 years.
3) The exchange only allows you to see what everyone else sees. No "peeking at other people's cards".
If they still do HFT with these conditions we might eventually see an improvement in algorithms, software quality and testing.
[1] Rollbacks are only allowed if it's not your fault e.g. the casino aka exchange screws up big time (slow downs don't count, going down doesn't count, exchange treating 1=2 counts, exchange treating buy as sell counts.).
Well, Thanks for Setting Me Straight (Score:5, Interesting)
Obviously I can't really elaborate on algorithms, but suffice it to say your understanding is naive at best - you're talking 2004 type games.
Great so you can't tell me why my understanding of how HFT works is wrong and I'm talking about "2004 type games" which would explain why I read about automated trading algorithms losing Knight Trading $440 million two months ago [slashdot.org]? Tell me, all those protection measures and penalties, did they protect the company running the automated trading software or the parties who engaged with trading with the automated trading software?
This stuff is almost always blown out of proportion and you'd never read about the actual workings of the regulation and clearing processes which protect all players
"Protect all players" you say? So that would mean that everyone gets paid when someone screws up big time? Well, I bet they're learning their lessons. I think what you mean is that it "protects the big firms that are doing the HFT" while the market is just a big massive beast ripe for the skimming?
And when trading systems do go awry most exchanges have built-in and often automated undo not to mention penalties.
So, when I buy stock in Wal-Mart and my "algorithm" (my brain) was screwed up, where's my automated undo button?
This is a game where someone's loss is almost always another party's gain. There is no way to "protect all parties involved" with that sort of game. It's the nature of the goddamn game.
If you're just some guy taking the highest paying programming job, I'm not mad at you -- that's capitalism. But if you're actually running the show or defending your boss, you and I are basically at polar opposites. HFT doesn't provide anything and receives an insane amount of cash. Betting on arbitrages isn't betting, you're basically taxing everyone else little bits of money and just being a huge fucking leech.
Re:Trading's Too Fast When It Ceases to Mean Anyth (Score:4, Interesting)
I see it as more of a high-end gambling. You're trying to make money, with money, hoping to outguess someone else. It's really not all that different than going to vegas and playing the poker table. If you can find a few suckers, and avoid being one yourself, you can make money.
All these people are doing is trying to speed up the process and make it easier to do. Some opportunities may only exist for a few seconds. Someone decides to sell when they shouldn't and causes price to dip momentarily. If you have automated systems in place you can take advantage of it better.
The problem of course is the computers can get really twitchy and cause shockwaves of trading to occur. My favorite thing to watch is how people scream about how all this money "just disappeared". Idiots. The money was never there to begin with. There was just an opportunity for money and now there isn't. Big difference. Money doesn't disappear, liquidity disappears. But that can cause problems for companies that are running closer to the edge. When there's a "crash", the money is just taken back out of the system. 100% of it CAME from the investors (the public) and for awhile some of it was in the hands of companies, selling stock is like getting a loan from the public. A crash just means people try to take their money back. They won't get it all back, the company has some of it and the public has some of it. No money disappears, you just don't see as much anymore.
It's all an illusion created by liquidity. It's like a group of three friends passing a $1 bill around. All three of them see $1 and get the impression that it looks like $3 since all three of them see $1. Until someone pockets the bill and the other two don't see it anymore, and start freaking out wondering where the other $2 went. A loss in liquidity makes people think (their) money has vanished, and someone must be to blame, and tends to cause a panic.
Same effect with stock. If you buy a stock from a company for $1, where's the money? They have $1 cash and you have $1 in stock. Does that mean that $2 exists in this closed system? NO. There was $1 and there still is only $1. If you turn around and sell your stock back to the company, you have $1 cash and the company, well their own stock isn't worth anything to them unless they can sell it so they really have nothing. So now we went magically from $2 to $1 as a dollar disappeared? NO.
Re:Trading's Too Fast When It Ceases to Mean Anyth (Score:5, Interesting)
It's more and more about what algorithms your "opponents" are using and what your algorithms are set at.
Only if you are in it for day-trading profits. And if you are, well, you deserve to be beaten senseless by some HFT algorithm.
If you're a long term investor, with a time horizon of many decades, this doesn't matter. For example, I have a stock I bought 15 years ago. It has gone up by around 3X in that time. HFT makes no difference to me when I've held a stock over many years or decades. The exact microsecond it sells doesn't matter to me after a period of decades.
The fundamentals are driven by stock valuations, which are based in what people guess about the future of the company. You can be as informed about that as anybody. If you believe a company will do well over the long haul, buy some of their stock. Don't worry about HFT. You don't have to microsecond-time your sale and beat some other HFT algorithm when you've made a lot of money over years, rather than little bit over seconds.
It's more and more about what algorithms your "opponents" are using and what your algorithms are set at.
Only if you are in it for day-trading profits. And if you are, well, you deserve to be beaten senseless by some HFT algorithm.
If you're a long term investor, with a time horizon of many decades, this doesn't matter. For example, I have a stock I bought 15 years ago. It has gone up by around 3X in that time. HFT makes no difference to me when I've held a stock over many years or decades. The exact microsecond it sells doesn't matter to me after a period of decades.
This is a good point. But then the question becomes: what good does HFT provide? A lot of smart people are sucked into that sector because of the money, where they are arguably causing a net loss to society by participating in an arms race that does not produce any real goods or services. Those smart people could in principle be contributing to research and development in areas that actually improve everybody's standard of living, such as medical research and robotics - or perhaps even in economics when it comes to analyzing long-term successes (after all, genuinely improving the capital allocation in the long-term could be beneficially to society, unlike the short-term gambling that is happening these days).
With that in mind, there needs to be a discussion on how best to disincentivize this kind of extremely short-term behavior, where it is via transaction fees or via trading on heartbeats.