Effect of automated trading on Markets

[wilburbear]

Quote from martys:

If automated trading volume keeps its pace... it will be more like machines against machines. :eek:

Very good observation. About 6 months ago, I met a guy coming form one of the automated trading firms. He was now programming and making his own models. Anyway, he said the automated firms were designing programs that would enter an order and also detect the presence of other automated trades, and adjust to these other programs. The machines are talking with each other in a rudimentary way. And that was six months ago. Who knows what they have now.

How any individual can make money in that environment is beyond me.

 

[OddTrader]

Perhaps it has been algorithms against algorithms for long time. That would be unchanged forever, probably.

 

[ratboy88]

Quote from wilburbear:

Very good observation. About 6 months ago, I met a guy coming form one of the automated trading firms. He was now programming and making his own models. Anyway, he said the automated firms were designing programs that would enter an order and also detect the presence of other automated trades, and adjust to these other programs. The machines are talking with each other in a rudimentary way. And that was six months ago. Who knows what they have now.

How any individual can make money in that environment is beyond me.

there are two problems i see with the automated players. first of all it is hard as hell to compete with someone that pays almost no commission fees. granted they have overhead but bottom line they are trading for pennies and thusly kill almost all good moves. secondly, they make it very difficult for the specialist to run his book. when the auto's step up volume the specialist has to let the auto bidding take over.....he will arbitrarily freeze the book and when he is finished filling orders he puts it back on autobid and lets it chop the hell out of everybody. in conclusion, i don't feel automation is compatible with the specialist system and needs to be one or the other. if they go the "all electronic" route then they have got to increase the minimum price increment due to fragmentation. there is no way this fragmentation is helping joe six pack investor whether he invests on his own or participates via an institution. the fragmentation is only benefiting the biggest auto players such as Goldman and JPM.

 

[wilburbear]

I heard hedge funds that go long and short stocks, are, on average, down on the year. Machines had to have contributed to that. Can anyone confirm that about hedge funds? If true, money will come roaring back out of hedge funds and into simple old t-bills which are much safer, less volatile, and yield hundreds of basis points more.

These are astonishing times. A journalist recently postulated "the end of history", as we would now all be democratic and capitalist. Is it the end of equity trading as we know it, when even hedge funds can no longer make money trading stocks? You will never be able to go back to the pre-machine days.

 

[nugundam]

Quote from ratboy88:

there are two problems i see with the automated players. first of all it is hard as hell to compete with someone that pays almost no commission fees. granted they have overhead but bottom line they are trading for pennies ....

-----------------
Very interesting observation. But thats why I think we're playing in a much more level field now with these new prop. trading firms where you practically pay zero comissions (swifttrade for example).

 

[FuturesTrader71]

What I find interesting is the possibility that some virus, nasty geopolitical event (dirty bomb) or whatever causes massive movement will cause these programs to puke out and basically result in a similar trade as a fat-finger. News events are probably not built into these algorithms. An exchange freeze or a real threat or disaster could easily put a few clearing firms/ banks under.

Unfortunately, it takes a disaster to cause changes that would manage or prevent this and I'm not sure how many out there have it built into their system. In a way, it would make some sense to create products that are more machine friendly and others where machines would be prohibited . For example, a DOW, a mini-DOW, and an e-DOW. Or Russell 1000, Russell 2000 and e-Russell.

I know... I'm dreaming...

 

[Trader5287]

Quote from nononsense:

Is that so? How do you know this?

Load the next month on your trading screen.

Thus in cac40 you would have May and June up and DAX Jun and Sept. Mine comes via DTB.

Now look at that volume on the bid and ask - notice anything?

Geo.

 

[FredBloggs]

Quote from FuturesTrader71:

What I find interesting is the possibility that some virus, nasty geopolitical event (dirty bomb) or whatever causes massive movement will cause these programs to puke out and basically result in a similar trade as a fat-finger. News events are probably not built into these algorithms. An exchange freeze or a real threat or disaster could easily put a few clearing firms/ banks under.

Unfortunately, it takes a disaster to cause changes that would manage or prevent this and I'm not sure how many out there have it built into their system. In a way, it would make some sense to create products that are more machine friendly and others where machines would be prohibited . For example, a DOW, a mini-DOW, and an e-DOW. Or Russell 1000, Russell 2000 and e-Russell.

I know... I'm dreaming...

exactly.

firstly i dont see what costs have to do with it. an investment bank or prop house will ALWAYS have next to no costs. so no change there. anyway, what difference does it make to your trading plan? what someone else pays has nothing to do with it. all it will do is encourage machines to scalp - thus providing liquidity. thats cool in my book.

as far as eye can see, these things come in lots of different flavors and colors. some algorithms scalp, some arb, some spread, etc. they just help with liquidity.

as ft71 points out, these machines have their downsides as well as the upsides. it is no panacea.

the main issue that i guess people moan about is the lack of trend. this has little to do with machines per se, but more to do with the prime brokerage model employed at institutions (imo). yes machines do play a big part in this model, but the main role is for efficient execution for huge positions - often with a long term view (still), not to shaft mr retail.

machines can never compute as many facts and information as the human brain. so give the human brain some discipline and the human brain will have the edge all the time.

so, maybe we already have an edge. seems like a lot of folk havent figured this one out yet and would rather develop a negative mind set about the markets/machines/cost structures/whatever are out to get them.

 

[EricP]

Quote from FredBloggs:

machines can never compute as many facts and information as the human brain. so give the human brain some discipline and the human brain will have the edge all the time.

I'm sure that Garry Kasparov would have strongly agreed with you, until he got his butt kicked by Deep Blue in 1997. In that matchup, you are not just talking about the discipline of the human brain, but you are talking about the undisputed #1 human brain in the game of chess. And still, given this apparent 'mismatch,' the computer beat him.

We shouldn't underestimate the ability of computers, in trading or other fields. Computers are not merely silicon devices tinkering brainlessly into the markets, but instead they are <b>tools</b> developed and implemented by some of the smartest brains in the industry. Computers become, in effect, an extension of the brain of the trader that programmed it, a very powerful combination, IMO.

-Eric

 

[nononsense]

Quote from FuturesTrader71:

What I find interesting is the possibility that some virus, nasty geopolitical event (dirty bomb) or whatever causes massive movement will cause these programs to puke out and basically result in a similar trade as a fat-finger. News events are probably not built into these algorithms. An exchange freeze or a real threat or disaster could easily put a few clearing firms/ banks under.

Unfortunately, it takes a disaster to cause changes that would manage or prevent this and I'm not sure how many out there have it built into their system. In a way, it would make some sense to create products that are more machine friendly and others where machines would be prohibited . For example, a DOW, a mini-DOW, and an e-DOW. Or Russell 1000, Russell 2000 and e-Russell.

I know... I'm dreaming...

Interesting point Trader71.
Is this realistic? You first would have to find out how many traders use identical automatic trading software in order to make this scenario realistic. I kind of exclude the erroneously called 'API trading gadgetry' as any serious tools. This segment only represents IMHO a continuation of old-fashioned dumb money.
I doubt that any serious automated market participant would likely share software with anybody. If it works for him, he certainly won't sell it. This is a theme that has been discussed over and over again.
Nevertheless, your question is a valid one. Ever since the introduction of the 200day MA, the markets have been affected by what all came after it, no doubt about this. Panics (black swans ) will keep on occuring. Hard to speculate about what people have built in as protection in their automation.

I am not sure whether prohibiting machines would be possible at all. Also, not sure about creating 'machine friendly products'

Be good,
nononsense