| | | That was a mistaken second negative: <You have a double negative about me reading about flash crashes. Are you saying that I only read about flash crashes from you? If so, you’re 100% wrong. Are you saying that someone discovered that flash crashes are not caused by HFT and how their trading system is set up? I doubt that’s true, but send a URL and I’ll read through the evidence.> Yes, I'm saying you only read about the CAUSE of flash crashes from me. The correct cause. Of course there were acres of writing about flash crashes in general and faulty ideas about the causes.
Yes, I'm saying the flash crashes are deliberate strategies, not "how their trading system is set up" in the sense that they made a mistake.
<OK, my turn: flash crashes are inherent signs of instability by HFT algorithms and provide no public service. > They are not intended to provide a public service. They are intended to provide a private profit. But they do provide a public service in that they hunt down unstable margined and Stop Loss positions and by eliminating such risk, make markets more stable.
The flash crash indicates that there was some instability but the purpose of a flash crash is to remove the instability by eliminating those unstable positions for the profit of the flash crash champions.
That will just make more paperwork and give unearned income to whoever charges the fee: <There should be a fee associated with every order entered into the market that is large enough to discourage HFT, but small enough to be absorbed by smaller traders. I assert that this will minimize or eliminate flash crashes> Neither would it solve the "problem" of flash crashes which are not actually problems. They are solutions.
I did not claim that: <Yes, but this does not support your assertion that flash crashes and longer term crashes are the same. > SOME longer cycle "flash" crashes are due to dueling computer systems versus human responses. But some are just responses to macro economic conditions. For example in 1929 there were no computers, but they had a crash anyway. Same in 1974. In 1987 there were computers getting into the act. In 1998 there was DE Shaw and Long Term Capital Management and computers were in business in a big way. In 2008 computers were big time and taking a full part in the process. Then there are crashes such as Blackberry, Nokia, L M Ericsson etc which are just companies going bust because of bad management using obsolete technology. Their crashes are not due to dueling computers.
Flash crashes are when there is a cycle down, then back up to more or less the same level. Those are due to:
On the way down:
Initial selling by Killer Whale HFT million mile a minute supersonic computers
Margin calls
Stop Loss orders
Dueling computers playing chicken
Panic by some computers
Panic by some people
Orders to sell at the "market price" [which turns out to be really low because the HFTs have cleaned out the rest]
Then, on the way back up:
Champion HFT MMaMSupersonic computer buying all on offer at the bottom
Second rate half a million mile a minute subsonic computers buying what they can at rapidly rising prices
Some humans who were quick or lucky getting a piece of the action
It doesn't matter than the sold Stop Loss shares don't come back into the market because the value of the shares is what it was, irrespective of who owned them. So the HFT winner simply keeps them as paper profit having bought them at the cheap Stop Loss price.
That's what really happens. It's not an accident of bad HFT design requiring a government department or arbitrary fees for putting orders in to buy or sell. Your "solution" will make things worse. As did the cancelation of orders [reduced trust in the market] and other proposed "solutions" to a non-problem. The best solution is to have buy orders ready to go when a flash crash happens.
Yes, you are right that there is a difference between observation and a theory. But you are wrong that a theory can't exist until there's an equation. Mathematics is just reasoning by another name. Reasoning can be done without mathematics. A dog can do observation, run and catch a ball. That doesn't mean it has done much reasoning and it has certainly not done maths. You might like to read about the definition of "theory" here: dictionary.reference.com Theory does not equal mathematics.
Darwin's theory of evolution didn't involve equations.
Spherical cow mathematics is inaccurate. <I am not surprised by this. You have many observations, but the problem is that you can’t determine the consistency of these observations without having a (real) theory to test them against.> Of course there's no equation for it, just as there is no equation for climate change. Some things are too hard to make models for them. Which is why the HFT million mile a minute supersonic computer people hire people with PhDs in mathematics to try to improve their modeling of what the heck will happen. Its not surprising that you are not surprised that I have not got such an equation. But that doesn't mean I have not got a theory. I have a theory - explained here [and over the last several years]. Flash crashes are not a glitch in the matrix.
And there won't be consistency in observations because the HFT computer reactions are not going to be consistent [as is obvious if you think about it for a moment]. You are imagining a dumb Newtonian pricing theory to be tested. But of course that's not going to exist. That's why there are PhD mathematicians dueling. Any theory would need to deal with interactive consciousness - adapting to reality as it happens and changing reality before it happens. Quantum mechanics and quantum computing is where it's going. "Head and Shoulders" theories are passe.
Mqurice |
|
