Skip to main content

Author Archives: Margo Reyfman

Algorithmic Trading

Argo Algorithmic Trading offerings include Argo Robots  – server-based automatic trading strategy hosting system and High Frequency Trading (HFT) framework.  Both solutions are seamlessly integrated with Argo Trading Platform components – Argo Trader, Order Router, MD Feeder and Risk Manager Server. Designer of automatic strategy  (for HFT or MFT) writes his code using our powerful ATP Trading API. API includes a complete set of order management, market data interfaces as well as a number of mathematical and statistical functions that are necessary for technical analysis based algorithms. The strategy is built into a dynamic link library (we provide examples) and injected into the Argo Robot Server. Here is Argo Robot Developer’s Guide.

Argo Algorithmic Trading  framework is written on C++ for optimal performance and portability. Source code licenses are available.

For web-based front-end developers, we provide RESTful Trading API.

Please contact us for information regarding new ATP trading strategy backtesting facilities.

Argo Robot Control Facilities

In Argo Robots environment a trader starts a new trading strategy instance using Argo Trader robot control panel.  Robot’s start-up parameters are submitted to Argo Robots  hosting server. The hosting server starts the strategy process (a robot). Robot receives market data, risk control messages and sends orders to the markets. It reports its status back to Argo Trader. For each running robot Argo Trader displays positions, P&L and, also, robot-specific information. Trader can control robot execution, stop and restart robot by sending commands and modifying robot parameters — on the fly!

Robot control facilities can also be integrated into your own application using our Argo Robot Control API.

Backtesting and Forward Testing

Argo offers a robust and functionally rich backtesting and forward testing facilities.

Our forward testing facility is based on market simulation module of the Argo Matching Engine. This infrastructure, which includes both backtesting and forward testing components, enables traders to rigorously test and refine their algorithms under realistic market conditions, ensuring optimal performance and reliability.

Custom Automatic Trading Development

We have accumulated a wealth of experience building automatic trading and market making strategies using our powerful Trading API and HFT framework. Please send us a message to inquire about our custom development services and provide contact information. We will get back to you promptly.

AI-driven Trading

We are also working on an AI-driven trading framework that applies supervised learning for market regime classification, ranks strategy candidates based on historical performance metrics, and uses Retrieval-Augmented Generation (RAG) with a fine-tuned Code LLaMA model to auto-generate ATP-compatible C++ strategy code. We plan to release the first version by the third quarter of 2025.

Argo Matching Engine

Argo Matching Engine is a highly scalable, low-latency order matching server — the software core of any electronic marketplace. It supports price-time priority and multiple variations of pro-rata order matching, with industry-standard FIX interfaces for order management and FIX/FAST over UDP multicast for market data. The engine also includes an implied matching module, and uses an RDBMS-backed order history storage system for full auditability.

The engine is built in modern C++, and runs on Linux and Windows. It can be ported to other major UNIX platforms upon request. A source code license is available, offering full transparency and customization flexibility.

Use Cases

Regulated Exchanges & ATSs
Deploy Argo as the core matching engine for equity, digital asset, or fixed income markets. Supports multi-venue routing, ATS-N compliance workflows, and integrations with RMS/OMS stacks.

Internalization & Off-Exchange Matching
Broker-dealers or liquidity providers can use Argo to build internal dark pools, improve quote internalization, or reduce market impact with smart matching logic and trade handling.

Commodity & Niche Marketplaces
Ideal for startups or regional platforms in metals, carbon credits, agri-commodities, and renewable energy credits. Includes auction support and customizable lot definitions.

Tokenized Asset Platforms
Use Argo to power order books for real-world asset token trading — such as real estate, gold, or collectibles — with fractional trading and custom matching logic.

Proprietary Trading Firms / Simulated Markets
Run low-latency simulations, backtests, or internal strategy competitions. Argo’s performance and control make it ideal for sandboxing or prototyping alpha strategies.

In-Game Economies / Digital Goods Markets
Gamified marketplaces benefit from Argo’s high throughput and API-first architecture to match items, skins, or NFTs in real-time with CLOB-like logic.

 

Key Features

  • Price discovery via continuous order matching  sessions, call auctions and IOI/RFQ dataflow;
  • Price-time priority and several variations of pro rata order matching algorithms;
  • Extendable set of instruments: equities, futures, options, bonds, FX spot, cryptocurrency contracts, forwards, exchange traded strategies (like calendar spreads, butterflies), interest rate swaps, credit default swaps;
  • Support of dynamic intra-day instrument creation;
  • Implied engine: implied IN and implied OUT;
  • Order Types: Market, Limit, Stop Loss, Stop Limit, MOO, MOC, Iceberg;
  • Time-in-force options (TIF) and execution instructions: Day, GTD, GTC, GFS, FOK, IOC, AON;
  • One-cancel-the-other (OCO) and if-done combinations;
  • Validation of order price alignment, price range, minimum order size, order size increment, and more;
  • FIX 4.2 and FIX 4.4 interfaces for order management;
  • Market data publishing: BBO, last trade, trading session statistics, aggregated market depth (full snapshot, incremental updates), market-bo-order, instrument definitions;
  • Support for trading session scheduling on market, channel, and instrument level;
  • Administrative HTML and command line interfaces provide trader session statistic, trader activity, order book, fills, and market depth views;
  • Support for “manual” matching for dealer-desk setups via IOI/RFQ dataflow;
  • Scalability: advanced multithreading, utilizing multiple processors efficiently,  deployment of multiple matching engines  in one trading environment;
  • Fractional matching;
  • Seamless integration with Argo Trading Platform components: ATP Order Management System, Market Data Feeder, Risk Management Server, and Argo Trader front-end;
  • Trade capture FIX session;
  • Integration with trade reporting facilities (like FINRA TRF).

Market Simulation

Argo Matching Engine can be used as market simulation system . To simulate market we feed server by real-time or predefined market data.

See how Argo Matching Engine has been used to build a powerful CME Simulation system.

Matching Engine Source Code License

Argo Matching Engine is a White Label product. You can purchase the matching engine source code by paying one-time license fee, and distribute the application to your clients without paying additional fees. Argo provides support, training and custom software development services.

Contact us to learn more about Argo products and services.

Simulate Market with Argo Matching Engine

Developers of trading applications normally require special order management and market data test connection or paper trading environment to utilize forward testing scenarios. Although many exchanges provide market simulation, real life market liquidity is not always present. Diagram above shows how we have used Argo Matching Engine to build a powerful simulation environment for CME. Now our customers who are writing automatic trading strategies with Argo Trading Platform and Argo Trading API can get the necessary mileage before going into production. To assist in realistic testing of automatic trading applications we have added a simulation of different network conditions and timings.

Simulate Market with Argo Matching Engine

How Market Simulation Works

We have substituted Argo Matching Engine inbound FIX interface with CME iLink complaint module and used CME MDP 3.0 and FIX/FAST 2.0 protocols to publish market data. We have designed a special Argo Matching Engine plug-in module, Order Flow Simulation (OFS). OFS is responsible for converting market data coming from the exchange into buy and sell limit orders that are posted into Argo Matching Engine Order Book. Client orders are matched against OFS contra orders. If trade is made, the corresponding Execution Reports are sent back to the owners’ FIX sessions and OFS module gets notified as well. OFS keeps track of its orders in Matching Engine Order Book and adjusts total quantities on each price level as necessary, submitting new orders and/or canceling existing ones based on new market depth updates coming from the exchange.

Argo Matching Engine can be configured to load several OFS modules to simulate different markets. OFSs that can capture CME, ICE, MOEX, BitMEX are available.

Free One-Hour Design Session!

Please contact Argo SE to find out how you can use Argo Matching Engine to simulate your market.
Allow one of our experienced software engineers to discuss the project requirements with you or your team, and produce a candidate system architecture diagram along with a draft project plan.

Argo C++ Trading API

Argo Trading API is a tool of algorithmic trading, high-frequency trading specialists and trading front-end application developers.  API provides order and quote management,  market data, risk information and technical analysis interfaces. Click on the class diagram to get more details, or see complete API reference.

API-based algorithmic trading strategy runs  on the server side. It is seamlessly integrated with ATP Order Management System,  Market Data Distribution facility, Risk Management System and (optionally) reports its state to ArgoTrader front-end.

Argo Web API

Argo Web API is built as a set of RESTful request/reply and WebSocket data streaming services. It can be divided into three sub-domains: order management, market data and risk information. An authorization (logon) procedure is required before any Trading API calls can be made. Typically, Web API returns data in xml format. Requests are HTTP Get and Post messages. Some requests contain JSON formatted objects.
Read documentation for Argo Web API’s:

Argo High-Frequency Trading Solution (HFT)

Argo HFT integrates Trading API, ultra-fast Order Management System and Market Data Distribution facility with in-process trading venue order management and market data  adapters. The solution comes in one lightweight Linux package with network interface cards by Solarflare and OpenOnload kernel-bypass IP stack.

Backtesting and Forward Testing

Backtesting allows an automatic strategy developer to verify his algorithms implemented in API-based application (robot) without sending orders to live markets.

ATP Backtesting Workflow:

  • Collect market data
  • Configure API-based algorithmic trading application to run in market simulation mode
  • Insure output of strategy performance metrics, P&L
  • Run API-based program(robot) in simulation mode
  • Analyze performance metrics, adjust algorithm, repeat

ATP Forward Testing:

Traders often make the mistake of relying entirely on backtesting results to determine whether the system will be profitable. While backtesting provides traders with valuable information it is only one part of the evaluation process.

Forward performance testing, also known as paper trading,  is a simulation of actual trading and involves following the system’s logic in a live market. All trades are executed on ATP Matching Engine that runs in market simulation mode, no orders are sent to real markets.  Argo Matching converting market data coming from the exchange into buy and sell limit orders that are posted into  its own Order Book. Client orders are matched against simulated contra orders. Simulation logic  adjusts simulated orders quantities on each price level as necessary, submitting new orders and/or canceling existing ones based on new market depth updates coming from the exchange.

API-class-diagram

News:

read all news»
Email