Archive for Technology

Jan
13

Multi-Threaded Kernel Technology

Posted by: Lawrence Chan | Comments Comments Off

Are you having problems with:

  • Dropped ticks
  • Delayed ticks
  • Crashes
  • Lock ups
  • Top-of-the-line computer, mediocre performance

And the problems worsen when you track more symbols? It’s likely that the problems are caused by incompatibilities among your trading platform and your computer.

The Cause

Charting programs originated in 2000s or earlier, were written with old computers in mind. Because personal computers at that era were not good at multi-tasking, early charting programs are single threaded. These programs worked okay in that era because computerized trading was not wide spread and market activities were low.

Fast forward to present. Most traders now trade with computers and specifically, through the Internet. Analysis is done faster. Orders are placed faster. This creates a surge in trading volume. It’s now common for an active stock to generate half million trades per day. Such high trading volume is unprecedented just a few years ago.

Old charting programs, entry-level applications, and bundled solutions are not prepared to handle such extreme volume, period. Problems surface when you needed your application most, like during the time when the market activities is reaching its peak. The symptoms are clear – single threaded (and most badly designed) programs will drop and/or delay ticks. Some even crash.

Buying a faster computer does not solve the problem. The inefficiencies are there within those trading applications, not the computer.

NeoTicker® Advantage

NeoTicker® is a modern multithreaded application designed to handle intensive data load in real-time by fully utilizing the latest CPU technologies. It is created from ground up for real-time intraday charting and data processing.

Most of its peers are developed for end-of-day usage and later modified into processing real-time data, hoping that their applications would still hold up. Existence of such software is an insult to the intelligence of the general public and irresponsibilities of the developers.

To process data the right way, workload has to be spreaded across multiple threads. Each thread should be responsible for a relatively simple task. Because the tasks are simple, the threads practically run simultaneously in real-time.

For effective real-time performance, NeoTicker® has proprietary load balancing technology to ensure every thread is well behaved and consume computer resources in a reasonable manner.

The result is ultra smooth real-time performance only NeoTicker® can deliver.

multi-threadReduce Data Delay

NeoTicker® uses separate threads to handle distribution of ticks.

After the receiving thread receives a tick, the distribution thread takes over almost instantaneously to distribute the tick to individual windows.

Because the time spent on distribution of data is minimized, ticks are rarely delayed, even in the busiest market.

Data Robustness

For each thread, there is a dynamically adjusted queue for buffering data. If the computer is too busy to handle a tick, the tick will be stored in the queue and no data will be lost

Effective Use of Your Hardware Investments

You don’t need a high-end computer to run NeoTicker®. NeoTicker® is very efficient. For average users, a mid-range computer is more than sufficient.

If you require heavy computation, you will be gladed that NeoTicker® can effectively utilize the latest hardware. NeoTicker®’s multithreaded architecture fully utilizes the hardware that is built-on multiple virtual or physical CPU’s. There is a marked performance improvements when you run NeoTicker® on multiple core computers.

Categories : Technology
Comments Comments Off
Jan
13

Indicator/Data Logistic Object Model

Posted by: Lawrence Chan | Comments Comments Off

programminglevels

Categories : Technology
Comments Comments Off
Jan
13

Transaction-Oriented Model Technology

Posted by: Lawrence Chan | Comments Comments Off

NeoTicker® implements trading models in a formal manner. A unified framework handles all trading related activities – trade simulation, trading systems, brokerage interface. This unified framework gives you the ideal balance between ease of use, flexibility and consistency.

Framework Overview

The framework on trading models is divided into three logical elements. All three elements are linked by a common real-time database:


transactionoriented_large

  • Order issuers, e.g. manual order entries, trading systems
  • Order processor, e.g. NeoTicker® Order Processing Engine, interface to real-life broker
  • Queries – system performance viewer, trading systems that rely on results of other trading system
    The database is responsible for storing orders, transactions and positions. The database makes sophisticated queries into trading history possible.




Trading Systems

Simply put, NeoTicker®’s trading system architecture is the most powerful and flexible one in the market today. No other software gives you the combined power and consistency NeoTicker® offers.

NeoTicker® is designed for real-time deployment. Your trading systems will run with the same level of performance in real-time as in historical testing.

In summary, NeoTicker®’s trading system architecture gives you the following benefits:

  • Based on our unique indicator framework – trading systems have access to advanced indicator features such as compress series, tick update, indicator-on-indicator and more
  • You have a variety of ways to implement trading systems, from Backtest EZ where no programming is required, to easy-to-use formulas, to flexible scripting languages and external programs though IDL interface
  • At the point evaluation – it is impossible to accidentally peek into the future and create an unsound trading system
  • Portfolio support for all order commands without explicit symbol switching
  • Query into database to implement order management and dynamically adjust trading systems’ behavior based on trading history

Order Processing Engine

NeoTicker®’s Order Processing Engine processes orders from Trade Simulator and trading systems. The engine supports many order types and provides realistic order filling at tick level.

Order processing generates transactions, which are logged into database so you can easily evaluate trading performance.

Benefits:

  • Maximum consistency between orders filled by different parts of trading framework
  • Tick level processing for historical and real-time tick data
  • Order types supported – fill or kill, day, good ’til cancel
  • Tick level fill type – fill by tick, fill by bid/ask
  • Non-tick level fill types – exact, average or worst price

API for Brokerage Interface

Our team works with brokers to provide interface to place orders from NeoTicker®.

When brokerage interface is enabled, orders are routed to the broker for processing. If the broker provides order status, you have the choice to use the broker’s order status instead of Order Processing Engine for actual filled prices.

System Performance Viewer

System Performance Viewer utilizes the database to query historical trading performance. Because the database logs all orders, transactions and positions, sophisticated query is possible. System Performance Viewer can produce comprehensive trading reports.

Here is a sample list of reports available in System Performance Viewer:

  • Trading summaries for everything in portfolio
  • Order list and distribution
  • Transaction list and trade price distribution
  • Position list and statistics
  • Runs and equity change distributions
  • User defined custom statistics
Categories : Technology
Comments Comments Off
Nov
22

Super Position Technology

Posted by: Lawrence Chan | Comments Comments Off

Time series transformation theory meets with non-conventional charting styles

The timeless Point and figure charting, the Japanese Renko or Three Line Break, are all excellent concepts that transform the underlying price data information into something that encapsulate specific principles in price movements, giving the user of such charting style a different perspective into the price history.

At TickQuest, we successfully modernized and unified the concept of time series transformations into a programming framework where users can easily implement various kinds of transformations, from the classic ones, to highly customized ones designed by the users.

Promotes Consistencies in Transformation Rules

Most charting platforms do not handle non-conventional charting style consistently over historical data and real-time updates. Designers and programmers of these platforms do not have the expertise in financial data processing, thus failing to understand the importance of consistencies in generating these charts is very important.

Super position technology formalizes the transformation rules, making it possible to generate good consistent specialize charting (i.e point and figure, renko, three line break, volume range, momentum bar, etc.) based on a base data series specification, across both historical data and real-time updates.

Promotes Innovations thru Plug-in Architecture

Super position is extensible through plug-ins created by users or third party developers. Researchers can now easily specify transformation rules that can be implemented into programming code for practical usage without ambiguity. Potentially better ways in data representation can be researched and developed in the future at a much faster pace.

Improves Consistencies in Trading Model Backtesting

Specialized charting has always suffered in terms of trading model development. Trading rules have never been easily defined for non-fundamental charting styles. For example, a point and figure chart cannot be properly backtested with rules based on its bars only because too much information regarding the underlying price series is missing. Orders just cannot be filled at the price you think possible with the point and figure bars.

With super position technology, system developers can now easily develop trading models to overcome this kind of issues and backtest with confidence that they can get consistencies on any kind of custom charting styles in their backtesting process.

Categories : Technology
Comments Comments Off
Nov
22

Grid Optimization

Posted by: Lawrence Chan | Comments Comments Off
Categories : Technology
Comments Comments Off
Nov
22

Environment Simulation Technology

Posted by: Lawrence Chan | Comments Comments Off

Ever imagine that you can replay the markets in front of you tick-by-tick, in the complete real-time trading environment you have created or customized, as if you are replaying a video?

Our environment simulation technology can do exactly that, with advanced features allowing you to play, pause, fast forward market data as if you are replay a video. You can even replay at the smallest possible resolution of one tick at a time (just like replaying video frame by frame).

There are many exciting applications with this useful technology. Here are some examples.

Streamline the Trader Training Process

All forms of attempts to replay the markets exist. From trying to watch a historical chart and scrolling that manually bar by bar from left to right, to simple automation tools adding bars to a chart one bar at a time. Countless number of people has spent millions of hours doing just that, so that they can train themselves without the benefit of peeking into the future on a chart with all the outcomes already presented itself to them.

That helps but not quite enough. Historical chart on a single instrument does not reflect the proper environment, or the setting, when the price series was developed. Short of those information, you are justifying the price actions and behaviour on a chart as if that is the only factor that matters.

Maybe. Maybe not. If you do believe that markets are inter-related and that the scenario at the time an instrument was traded would determine the outcome with greater accuracy, then, replaying the history of a single instrument is not the answer.

Our environment simulation tackle this exact problem by replaying the complete market environment you deem important for your decision making. That means, you can see the markets replay themselves in front of you in concert, as if you are monitoring them in real-time.

Improve the Indicator Design and Development Cycle

When designing a new indicator, one of the major obstacles is making the indicator responsive at the right time under the right condition. Using purely historical data with OHLC bars, the actual real-time development of the indicator values are not available for the indicator designer to verify the stability of their works.

With our technology, indicator developers can replay the data one tick at a time, and check on the behaviour of their work easily.

Efficient Pre-Deployment Analysis of Trading Models

Before a trading model goes live, be that manual order entry or fully automated system, test drive time period has to be spent on monitoring the real-time behaviour of the model, before actual deployment takes place. During such time period, order placement expectations, special situation handling logic, etc. will be added to the model based on potential issues observed during the test drive.

With environment simulation, one can efficiently stress test a trading model, without waiting until the test drive time period to figure out the potential problematic situations. Better model can be built and shorten the time requirement for test drive.

Categories : Technology
Comments Comments Off
Nov
22

Tick Precise Technology

Posted by: Lawrence Chan | Comments Comments Off

Classic technical analysis emphazes on the use of indicators, drawing tools, etc. to highlight and summarize the information on a chart. Decision making becomes a simplified process of looking into these summaries, as oppose to the details that formed the price data series in the first place.

Tick Level Data Analysis is Now a Necessity

As intraday price actions and program trading activities keep increasing, traders can no longer depends on such tools only due to the inherited weaknesses in these tools as they lag behind the price movements significantly when a market goes wild.

Tick precise technology is invented to allow indicators and tools to be developed so that they can process tick level data and produce summary information that was not possible with just OHLC bars.

Historical Data Generation Made Easy

One of the most difficult tasks in making tick precise indicators is to generate the correct historical indicator values within a chart. Normal chart data are loaded using the user specified information and the data series themselves do not have the tick level information needed by tick precise indicators.

To resolve that you can conduct tick replay on a chart to allow the tick precise indicators to properly generate their historical values directly from the tick data. Historical data and real-time update results can archieve maximum consistencies through this method. Making it possible to design and see how such indicator works on historical data, not just updated on streaming data.

During replay, tick data files are read and ticks (i.e. individual trades and their respective bid and ask updates) are used to reconstruct historical bars, just like real-time. Indicators are triggered on a tick-by-tick basis for recalculation. With tick replay, you can have confidence that your indicators and trading systems will work as expected.

Advantages with Tick Replay on a Single Chart:

  • Tick replay is designed to play unlimited number of days of tick files. NeoTicker® carefully manages computer resources so it is possible to play over 1 month of tick data of multiple symbols into a chart
  • Tick replay is very fast. A single tick file with over half million ticks can be replayed within a minute
  • Tick replay does not interfere with regular real-time tick receiving and processing. You can use tick replay on isolated charts while monitoring the market in real-time
Categories : Technology
Comments Comments Off
Nov
22

NeoBreadth® Technology

Posted by: Lawrence Chan | Comments Comments Off

Revolutionize Market Internal Anaylsis

Classic market internal data like advance/decline issues, TICK index, TRIN index, etc., broadcasted by the exchanges, represent statistics on the broad market, not truly correlated to the specific market indices or segments (S&P500, Nasdaq 100, the mining sector, the financial sector, etc.) you are dealing with. Should someone interested in generating these kind of specialized statistics, the task is so tedious, making it difficult for further research advancement in this area.

NeoBreadth® is designed as the solution to this problem. It revolutionizes the whole concept of market internal analysis. Analysts and Market Technicians can now easily collect and generate specialized statistics on multiple baskets of symbols, in real-time. New ideas or research concepts can be done easily as NeoBreadth® can reconstruct the historical data on the special statistics you’ve just created.

Encapsulate Complex Market Dynamics with Compact Formulas

To create your own market statistics, you just need to focus on one thing – the data you are interested to collect and analyze. For example, to collect real-time advance issues data on S&P500, it takes just an one line formula, Last > PrevClose. The rest of the data collection process is then handled by NeoBreadth®.

More complex concepts, including ones you wanted to research into, can be done the same way. You can focus on the idea itself, as oppose to wasting time on worrying about low level issues from connection to data services, data file processing, etc.

Extreme Efficiencies

neobreadth_w320One of the biggest obstacle for real-time market breadth analysis is that the computation load can be stressful for your computer. Most trading platforms simply break down when you make them collecting one or two customized market internals over just a hundred symbols.

That is not a problem at all for NeoBreadth®. NeoBreadth® is part of NeoTicker, our real-time platform designed to handle computations on thousands of symbols efficiently. There is no parallel to this level of efficiencies at all from any other trading software available nowadays.

Versatile Platform Integrations

Data generated with NeoBreadth® can be used within other software easily. For example, you can pull the data out to be used within Excel, or in fact any other software capable of standard Microsoft Windows integration technologies.

Categories : Technology
Comments Comments Off