CQG API and MATLAB® FAQ

8x2870 or above.

Yes, that is possible using MATLAB in CQG.

There may be depending on the computer; however, CQG believes the best solution currently is the NVIDIA Tesla card. It can process anywhere from 10 to 50 times faster. It is the preferred option for performing complex tests in MATLAB. MathWorks supplies an add-on to MATLAB called the Parallel Computing Toolbox, which allows you to take full advantage of multi-core, and the MATLAB Distributed Computing Server, which supports grid architecture.

You can write an indicator of indicators in CQG or in MATLAB.

COM is generally faster than other interfaces and is not tied/limited to a specific technology (like DDE to Excel, etc.). We optimized our COM interfaces, so that they would not introduce latency and still be very flexible.

CQG fees for the API are determined by the level of service you require. Additional information is available. For MATLAB, the minimum API requirement is Base Historical Data API. API options include:

• Real-Time Data API:  $45/month
• Trading API:  $245/month (requires Standard or Advanced Trading)
• Base Historical Data API:  $100/month
• Extended Historical Data API:  $300/month
• Extended Historical Plus Data API:  $1,000/month
• Unlimited Historical Data API:  $6,000/month. A one-year prepayment is required.

The DOMTracker study tracks the size of resting orders in the four queues above the best ask and the four queues below the best bid. If the DOMTracker readings are low, then traders are either entering market orders or are on the sidelines.

To use MATLAB functions (canned or custom) within CQG, you do not need to code the API. For more advanced analysis within MATLAB, you will need to code the API. CQG provides the API samples of integration with MATLAB. You need to have some programming experience and knowledge of MATLAB in order to code the API effectively. CQG can only call one value from a MATLAB function at a time; therefore, more sophisticated analysis requires the use of the CQG API.

CQG API is a COM server. Our API samples illustrate its usage.

Data can be purchased from the CQG Data Factory™. If you are enabled for CQG Backtesting, then 18 months of 5-minute data and multiples thereof are available. Five years of data is available at an additional charge. In terms of utilizing the API for historical data, there are four options to choose from:

Data Types	API Historical	API Extended Historical	API Extended Historical Plus
Intraday (Greater than 60 min.), Daily, Weekly, Monthly:	300 Bars	1,000 Bars	4,000 Bars
Intraday (Less than or equal to 60 min.):	3,000 Bars	10,000 Bars	40,000 Bars
Tick Data:	2 Trading Days	2 Trading Days	2 Trading Days
CVB, P&F, and TFlow Bars:	5 Trading Days	10 Trading Days	20 Trading Days

There is unavoidable latency between the CQG API and MATLAB. However, with high-end computing power, this should be kept to a minimum. Given the fact that MATLAB runs in a different processor on a multicore system, CQG and MATLAB should have minimal impact on each other.

Yes. The data can be retrieved via the API.

Yes. With CQG, you can display market data in the form of constant volume bars (CVB), which build the bars based on a threshold of executed volume or tick volume, not time, as well as TFlow® charts. TFlow charts plot the best bid and best ask at the highs and lows and are color-coded based on the amount of executed volume. Visit our Web site for more information on CQG charting. MATLAB also supports a range of charting from simple line plots and bar charts in base MATLAB to more specific candlestick charts and cash flow plots (functionality within the Financial Toolbox).

DDE is generally old and heavy technology. Just changing the data source from DDE to API will improve the performance. In addition, if the calculations in question are complex, then MATLAB is a better place to do them than Excel.

MATLAB is more effective for development because it provides an extensive mathematical engine while C++ is a tool that allows you to create anything from scratch, which consumes resources.

You can call DLLs from MATLAB. Three possible techniques include the function loadlibrary, the MEX (MATLAB executable) interface, and the COM interface. CQG also provides the functionality to call DLL libraries for use within CQG.

If you write a MATLAB custom function, run it in MATLAB and MATLAB indicates any errors in your programming.

CQG supplies whatever market data MATLAB requires to perform the calculations. The user decides whether the data is stored in MATLAB or in the CQG API. MATLAB can also write data to Excel, databases, or standard formats such as text, csv, and binary.

MATLAB has over eight thousand functions built in and available for programming. Refer to MATLAB’s user guides for more details.

You need MATLAB to connect to CQG. MATLAB offers toolkits, such as the Fixed Income Toolkit, that provide specific financial functions. For example, the Fixed-Income Toolbox™ extends MATLAB by providing functions for fixed income modeling and analysis. The toolbox includes tools for determining the price, yield, and cash flow for many types of fixed income securities, including mortgage-backed securities, corporate bonds, treasury bonds, municipal bonds, certificates of deposit, and treasury bills. MATLAB can guide you through what is best for your needs.

No, they are not. This link takes you to a list of available modules.

This link takes you to a list of the exchanges CQG covers for energy, which includes all the major exchanges worldwide and some OTC data. In addition, MATLAB offers a solution for energy suppliers, producers, and traders.

For a complete listing of MATLAB products and pricing by region, visit http://www.mathworks.co.uk/store/default.do.