Showing posts with label algorithmic trading. Show all posts
Showing posts with label algorithmic trading. Show all posts
Sunday, 25 March 2007
Algorithmic-Trading blog created
In order to spare professional traders and software engineers working on algorithmic-trading from my mindless brain droppings about science and software in general I have created an algorithmic-trading blog. I hope this enhances the reading experience for all involved.
Tuesday, 20 March 2007
Algorithmic-Trading for Dummies
Will all the algorithmic-trading gurus; the so called boy wonders of Wall Street please get their act together?!
The problem is interface consistency. Sure, there is the FIX Protocol 'standard' that most algorithmic engines use. However, even FIX is stretched to its limit when dealing with algos.
So here's my suggestion to all of you who write algo engine interfaces... attend the FIX Protocol's Algorithmic Trading Working Group and figure this all out!
The diversity in describing an algo (FIX fields, values, etc) are costing all of us time and money. Time to read and understand these so called specifications and money to constantly keep changing our Order Management Systems (OMS) to keep up with your so called great ideas (like calling an order a 'day' order when it really is a 'good until cancel').
Is it really that hard to specify the fundamentals of trading:
Furthermore, quite honestly, algos are money loosing ventures except for folks like me... We have to keep changing software to interface your silly algo engines so we keep getting a paycheck! However, CFOs are none to happy when passed in the hall and thanked by the likes of me for paying me in ever increasing amounts to invest in money losing ventures like algorithmic trading.
I now return you to your regularly scheduled production outage...
The problem is interface consistency. Sure, there is the FIX Protocol 'standard' that most algorithmic engines use. However, even FIX is stretched to its limit when dealing with algos.
So here's my suggestion to all of you who write algo engine interfaces... attend the FIX Protocol's Algorithmic Trading Working Group and figure this all out!
The diversity in describing an algo (FIX fields, values, etc) are costing all of us time and money. Time to read and understand these so called specifications and money to constantly keep changing our Order Management Systems (OMS) to keep up with your so called great ideas (like calling an order a 'day' order when it really is a 'good until cancel').
Is it really that hard to specify the fundamentals of trading:
- price
- time
- volume
- aggressiveness
- strategy
Furthermore, quite honestly, algos are money loosing ventures except for folks like me... We have to keep changing software to interface your silly algo engines so we keep getting a paycheck! However, CFOs are none to happy when passed in the hall and thanked by the likes of me for paying me in ever increasing amounts to invest in money losing ventures like algorithmic trading.
I now return you to your regularly scheduled production outage...
Thursday, 8 March 2007
Algos Prohibited From Dark Book Access
Many predicted it, some applaud it, others loathe the decision, but let's face it the battle lines are being drawn between dark books and buy side algos. The latest to throw down the gauntlet is Pipeline Trading Systems. Prior to that ITG-POSIT put the kibosh on algos (later they reconsidered).
Who will win this battle has yet to be determined.
What are the driving forces?
"It's the bandwidth, Stupid!", to paraphrase from a past election.
It takes a significant amount of work to implement, deploy and maintain a dark book (or an open book for that matter). True, it takes a significant amount of work to implement an algorithmic trading platform. However, algos don't face the same rules and regulations (e.g. RegNMS) as do their sell side brethren.
Let's frame this in an old world example...
Imagine a few thousand customers streaming into and out of a 'brick and mortar' store all just 'looking' but not buying. It costs the store owner money to maintain the large sales staff and floor space to handle the customer flow on the hope that someone will actually purchase some goods. If the overwhelming percent of people in the store are just looking, it's a time and money drain driving up costs and driving down revenue.
It's the same with dark books. All those algos add overhead w/o generating revenue (trades).
It takes a serious investment in hardware (servers, routers, databases), and software to support all the buy side algos that are just looking.
Who will win this battle has yet to be determined.
What are the driving forces?
"It's the bandwidth, Stupid!", to paraphrase from a past election.
It takes a significant amount of work to implement, deploy and maintain a dark book (or an open book for that matter). True, it takes a significant amount of work to implement an algorithmic trading platform. However, algos don't face the same rules and regulations (e.g. RegNMS) as do their sell side brethren.
Let's frame this in an old world example...
Imagine a few thousand customers streaming into and out of a 'brick and mortar' store all just 'looking' but not buying. It costs the store owner money to maintain the large sales staff and floor space to handle the customer flow on the hope that someone will actually purchase some goods. If the overwhelming percent of people in the store are just looking, it's a time and money drain driving up costs and driving down revenue.
It's the same with dark books. All those algos add overhead w/o generating revenue (trades).
It takes a serious investment in hardware (servers, routers, databases), and software to support all the buy side algos that are just looking.
Friday, 2 March 2007
Algorithmic Trading Problems
My main concern, four years ago, when algorithmic trading became popular was that it is a feed-forward system. The algorithms respond, in near real-time, to changes in electronic signals (bid, offer, trades) hoping to move the market in a direction favorable to the algorithm's goal(s). The problems with feed-forward systems specifically and controller theory in general are well documented and one of the primary problems is the inability of these systems to respond to novel situations like what was experienced this week in the markets.
I have worked in the financial markets the past seven years, and before that spent nearly ten years in the high energy physics research environment of Fermilab. I was more than a bit concerned when I started seeing algorithmic trading specifications, typically using the FIX protocol, show up on my desk about two years ago. My concerns center around the general lack of training most programmers receive regarding loop control and controller theory.
Therefore, this week when the markets started to fluctuate, in novel ways, I was not surprised when technical snafus at NYSE, Dow Jones and Arca started the electronic trading version of a microphone placed too close to a speaker. I have to wonder, but will probably never know, what role algorithmic trading engines contributed to the inflated tick volumes observed this week?
I have worked in the financial markets the past seven years, and before that spent nearly ten years in the high energy physics research environment of Fermilab. I was more than a bit concerned when I started seeing algorithmic trading specifications, typically using the FIX protocol, show up on my desk about two years ago. My concerns center around the general lack of training most programmers receive regarding loop control and controller theory.
Therefore, this week when the markets started to fluctuate, in novel ways, I was not surprised when technical snafus at NYSE, Dow Jones and Arca started the electronic trading version of a microphone placed too close to a speaker. I have to wonder, but will probably never know, what role algorithmic trading engines contributed to the inflated tick volumes observed this week?
Subscribe to:
Posts (Atom)