Algorithmic trading execution methods

So, I've come across names for various program trade algorithms and I'd like to understand what they are about. I figure, if these algorithms are common enough to have names and platforms advertise supporting them, then they are worth trying to understand. I realize these are mostly for institutions to try to get the best prices on buying and selling large quantities for the best execution but I think it's likely good stuff for a Trader to understand for better insight into overall market mechanics. Here is a list of terms:

Geurilla (Credit Suisse)
Sniper (Credit Suisse)
Ambush (Banc of America Securities)
Razor (Banc of America Securities)
Cobra (Instinet)
Nighthawk (Instinet)
Dagger (Citigroup)
VWAP and TWAP (Credit Suisse) -- I think these are pretty straightforward
Inline - FlexTrade
Iceberg - FlexTrade
Auto Market-Making (from Portware, do we have a good idea what this does? how it manages inventory and risk?)

There are more. I know the idea is largely to find liquidity and execute without signaling intent - but can anyone shed some light on the mechanics of doing that?

terms mostly from:
http://advancedtrading.thewallstreetwiki.com/directories/directory-futures-algorithms.php
http://w4.stern.nyu.edu/news/news.cfm?doc_id=6789
 
Cost is a major consideration here - some of those products cost like $100k to own/lease....

Like flextrade, apama, streambase, KDB+ etc....

And you'll need a superpricey datafeed...

Then you need custom filters to filter the data, specialized programming for your strategy that queries either SQL or a proprietary DB...

Some of these query languages are to me, impossible to figure out - I'm referring to KDB+ here, which uses Q, a language made by KX systems that provides extremely high speed execution and querying of time critical trading data...
 
Quote from psytrade:

Cost is a major consideration here - some of those products cost like $100k to own/lease....

Like flextrade, apama, streambase, KDB+ etc....

And you'll need a superpricey datafeed...

Then you need custom filters to filter the data, specialized programming for your strategy that queries either SQL or a proprietary DB...

Some of these query languages are to me, impossible to figure out - I'm referring to KDB+ here, which uses Q, a language made by KX systems that provides extremely high speed execution and querying of time critical trading data...

I understand. I'm not thinking that I would use them, but I'd like to understand more about how they work in order to understand what the Big Boys are doing (as much as practical).

Q sounds interesting. Do you use these technologies regularly?
 
I use none of it. I just researched it all and found out I was either too broke or not smart enough to figure out their query language. But broke came first :)

I don't do anything intraday so its really not much use to me having these kind of trading avenues....
 
caem - I can tell you some about this tomorrow. Bloomberg's "B-Pipe" is $4,500/month for digital data- so not out of control for data if you are at this level.

Are you looking to study these or are you looking lease?
 
Quote from spreadem:

Somebody wake me when cost come down to 5K per month.

How much do you expect to pay when most won't touch ATS unless it makes $5k/day
 
Quote from WinstonTJ:

How much do you expect to pay when most won't touch ATS unless it makes $5k/day
Most of the arb strategies that used to make 5K/day are so crowded that the easy/guaranteed money trickles in.
 
Quote from spreadem:

Most of the arb strategies that used to make 5K/day are so crowded that the easy/guaranteed money trickles in.

what??? :confused:

crowded trades usually mean there is no more money - especially when talking arb. as far as I know there is no easy money in crowded arb trades...
 
Back
Top