Imported from previous forum
Hello all. I want to write and test algorithms for level-II quotes for options trading. Before anything else, I want some sample data, because I want to find out whether this is for me.
This should be easy for those who write their own code. Can you provide me some real data, and the schema to interpret it? At the moment I can get only simulated data (from marketcetera).
Thank you,
Anuj
Hello all. I want to write and test algorithms for level-II quotes for
options trading. Before anything else, I want some sample data, because
I want to find out whether this is for me.This should be easy for those who write their own code. Can you provide
me some real data, and the schema to interpret it? At the moment I can
get only simulated data (from marketcetera).Thank you, Anuj
Anuj,
Level II data would need to be licensed from the exchange or ECN that generates it.
What we have at FIX is schema to publish algorithmic trading order types. It’s used when you want to allow others to use your algorithmic trading strategies. It defines the specific parameters that must be specified, and how they must be sent to cause the strategy to execute. It has nothing to do with how the algorithmic trading strategy (sometimes called a “black box”) operates internally.
The FIX standard is designed to be an interface between the Sell-Side (creators of algos) and the Buy-Side (users who are not creators). It’s also very handy for quant shops to use internally - the internal gurus make up all sorts of new algorithmic trading strategies for the in-house proprietary traders to use. Each of these is expressed in this new standard XML format. The proprietary traders drop these xml files into their order management systems, and up pops the trade tickets, all tuned to that specific new algo strategy, complete with order entry workflow, help during the order build process, field by field level validation, and a final comprehensive set of rules for validation (across all fields) that may be run after all parameters have been populated (such as when the “send order” button has been pushed.) Quant shops can then offer their traders the full set of their in-house algo strategies right alongside the externally available strategies without any recoding of the OMS. It provides very good isolation between the order management system (with its human computer interface/GUI) and the actual execution systems.
You can see all the documentation plus numerous xml instance files describing various algos, from 15 different firms, by logging in on the FIX site (guests may register quickly and for free at: http://fixprotocol.org/register/ ). You then log in and go to: http://fixprotocol.org/working_groups/algowg/documents. From there you can access all the .xsd and .xml files used to define algos.
Just a quick note – the FIX standard expresses an algo strategy in XML. However, contained in that XML is the map between parameters on the trader’s screen and standard FIX Protocol (ascii tag=value) format to be placed “on the wire”. So there is no XML used in actually sending the order out of the OMS. All of that is standard FIX protocol (for example FIX 4.2 format).
Also, most market data (last sale, bids, offers, etc.) is not expressed in XML. XML is thought to be too heavy a solution for huge volumes of market data.
For some examples of data formats for market data take a look at some of the options here:
http://www.batstrading.com/subscriber_resources.php
Rick at
ITSdoc.org
[ original email was from Ameya Nisal - ameyanisal@gmail.com ]
Hi Anuj,
I was curious to know wether we can program this algorithm of level II boxes for all the exchanges, especially the Indian exchanges (BSE, NSE).
We are a consulting group in India, looking forward to provide FIX products to Indian investment hubs, hence we are running a leg for programming products compatible for Indian markets.
Thank you.
Regards,
Ameya Nisal
Hello all. I want to write and test algorithms for level-II quotes for
options trading. Before anything else, I want some sample data,
because I want to find out whether this is for me.This should be easy for those who write their own code. Can you
provide me some real data, and the schema to interpret it? At the
moment I can get only simulated data (from marketcetera).Thank you, Anuj
Anuj,
Level II data would need to be licensed from the exchange or ECN that
generates it.What we have at FIX is schema to publish algorithmic trading order
types. It’s used when you want to allow others to use your algorithmic
trading strategies. It defines the specific parameters that must be
specified, and how they must be sent to cause the strategy to execute.
It has nothing to do with how the algorithmic trading strategy
(sometimes called a “black box”) operates internally.The FIX standard is designed to be an interface between the Sell-Side
(creators of algos) and the Buy-Side (users who are not creators).
It’s also very handy for quant shops to use internally - the internal
gurus make up all sorts of new algorithmic trading strategies for the
in-house proprietary traders to use. Each of these is expressed in this
new standard XML format. The proprietary traders drop these xml files
into their order management systems, and up pops the trade tickets, all
tuned to that specific new algo strategy, complete with order entry
workflow, help during the order build process, field by field level
validation, and a final comprehensive set of rules for validation
(across all fields) that may be run after all parameters have been
populated (such as when the “send order” button has been pushed.)
Quant shops can then offer their traders the full set of their in-house
algo strategies right alongside the externally available strategies
without any recoding of the OMS. It provides very good isolation between
the order management system (with its human computer interface/GUI) and
the actual execution systems.You can see all the documentation plus numerous xml instance files
describing various algos, from 15 different firms, by logging in on the
FIX site (guests may register quickly and for free at:
http://fixprotocol.org/register/ ). You then log in and go to:
http://fixprotocol.org/working_groups/algowg/documents. From there you
can access all the .xsd and .xml files used to define algos.Just a quick note – the FIX standard expresses an algo strategy in
XML. However, contained in that XML is the map between parameters on the
trader’s screen and standard FIX Protocol (ascii tag=value) format to
be placed “on the wire”. So there is no XML used in actually sending
the order out of the OMS. All of that is standard FIX protocol (for
example FIX 4.2 format).Also, most market data (last sale, bids, offers, etc.) is not expressed
in XML. XML is thought to be too heavy a solution for huge volumes of
market data.For some examples of data formats for market data take a look at some of
the options here: http://www.batstrading.com/subscriber_resources.phpRick at ITSdoc.org
Sorry, can’t be of much help for Indian markets.
I’d say you contact the stock markets directly, ask them about the protocols they will allow.
For US markets, I’d advise that you use a product like marketcetera rather than code it from scratch.
Good luck. It is not as complicated as people are projecting it to be.
/
Hi Anuj,
I was curious to know wether we can program this algorithm of level II
boxes for all the exchanges, especially the Indian exchanges (BSE, NSE).We are a consulting group in India, looking forward to provide FIX
products to Indian investment hubs, hence we are running a leg for
programming products compatible for Indian markets. Thank you.Regards,
Ameya Nisal
Hello all. I want to write and test algorithms for level-II quotes
for options trading. Before anything else, I want some sample data,
because I want to find out whether this is for me.This should be easy for those who write their own code. Can you
provide me some real data, and the schema to interpret it? At the
moment I can get only simulated data (from marketcetera).Thank you, Anuj
Anuj,
Level II data would need to be licensed from the exchange or ECN that
generates it.What we have at FIX is schema to publish algorithmic trading order
types. It’s used when you want to allow others to use your
algorithmic trading strategies. It defines the specific parameters
that must be specified, and how they must be sent to cause the
strategy to execute. It has nothing to do with how the algorithmic
trading strategy (sometimes called a “black box”) operates
internally.The FIX standard is designed to be an interface between the Sell-Side
(creators of algos) and the Buy-Side (users who are not creators).
It’s also very handy for quant shops to use internally - the
internal gurus make up all sorts of new algorithmic trading strategies
for the in-house proprietary traders to use. Each of these is
expressed in this new standard XML format. The proprietary traders
drop these xml files into their order management systems, and up pops
the trade tickets, all tuned to that specific new algo strategy,
complete with order entry workflow, help during the order build
process, field by field level validation, and a final comprehensive
set of rules for validation (across all fields) that may be run after
all parameters have been populated (such as when the “send order”
button has been pushed.) Quant shops can then offer their traders the
full set of their in-house algo strategies right alongside the
externally available strategies without any recoding of the OMS. It
provides very good isolation between the order management system (with
its human computer interface/GUI) and the actual execution systems.You can see all the documentation plus numerous xml instance files
describing various algos, from 15 different firms, by logging in on
the FIX site (guests may register quickly and for free at:
http://fixprotocol.org/register/ ). You then log in and go to:
http://fixprotocol.org/working_groups/algowg/documents. From there you
can access all the .xsd and .xml files used to define algos.Just a quick note – the FIX standard expresses an algo strategy in
XML. However, contained in that XML is the map between parameters on
the trader’s screen and standard FIX Protocol (ascii tag=value)
format to be placed “on the wire”. So there is no XML used in
actually sending the order out of the OMS. All of that is standard FIX
protocol (for example FIX 4.2 format).Also, most market data (last sale, bids, offers, etc.) is not
expressed in XML. XML is thought to be too heavy a solution for huge
volumes of market data.For some examples of data formats for market data take a look at some
of the options here:
http://www.batstrading.com/subscriber_resources.phpRick at ITSdoc.org
Hi Ameya,
It is possible to program any algorithm for NSE or BSE but you would have to take permission from the exchange and other formalities to be done.
Regards,
Varun
Hi Anuj,
I was curious to know wether we can program this algorithm of level II
boxes for all the exchanges, especially the Indian exchanges (BSE, NSE).We are a consulting group in India, looking forward to provide FIX
products to Indian investment hubs, hence we are running a leg for
programming products compatible for Indian markets. Thank you.Regards,
Ameya Nisal
My understanding is program trading in India requires regulatory approval granted in advance. During registration, all the internal math/logic models must be fully exposed in painstaking detail. As a result, the market for algorithmic trading Indian markets tends to be underdeveloped.
There are however, sophisticated algorithmic trading development groups in India building models for proprietary use in non-Indian markets. Also, there are plenty of top notch Indian born programmers working outside India to develop algorithmic trading techniques for deployment in various world markets.
Were India to modify its disclosure laws, my guess is the market opportunity would be very quickly addressed by several local and international algo experts.
Hi Ameya, It is possible to program any algorithm for NSE or BSE but
you would have to take permission from the exchange and other
formalities to be done.Regards,
Varun
Hi Anuj,
I was curious to know wether we can program this algorithm of level
II boxes for all the exchanges, especially the Indian exchanges
(BSE, NSE).We are a consulting group in India, looking forward to provide FIX
products to Indian investment hubs, hence we are running a leg for
programming products compatible for Indian markets. Thank you.Regards,
Ameya Nisal
I think even the DMA also has a lot of regulatory approvals in India, so if at all you want to use Algo Strategies for DMA orders, you may have to expose all your strat logics to the regulatory body along with your DMA platform details…a recent notification from BSE —http://bseindia.com/dma.asp
My understanding is program trading in India requires regulatory
approval granted in advance. During registration, all the internal
math/logic models must be fully exposed in painstaking detail. As a
result, the market for algorithmic trading Indian markets tends to be
underdeveloped.There are however, sophisticated algorithmic trading development groups
in India building models for proprietary use in non-Indian markets.
Also, there are plenty of top notch Indian born programmers working
outside India to develop algorithmic trading techniques for deployment
in various world markets.Were India to modify its disclosure laws, my guess is the market
opportunity would be very quickly addressed by several local and
international algo experts.Hi Ameya, It is possible to program any algorithm for NSE or BSE but
you would have to take permission from the exchange and other
formalities to be done.Regards,
Varun
Hi Anuj,
I was curious to know wether we can program this algorithm of level
II boxes for all the exchanges, especially the Indian exchanges
(BSE, NSE).We are a consulting group in India, looking forward to provide FIX
products to Indian investment hubs, hence we are running a leg for
programming products compatible for Indian markets. Thank you.Regards,
Ameya Nisal