Algorithmic Trading Overview

Introduction

• Diverse audience with varying experience in markets: Forex, options, and no trading experience.
• Discussion on the high cost of trading mistakes.
• Focus on algorithmic trading with a programming approach.

Key Problems in Algorithmic Trading

Data Acquisition

• Importance of data quality: Poor data leads to poor analysis.
• Free data sources like Yahoo Finance: Low frequency and indices; limited for retail traders.
• Paid data: Expensive and usually meant for institutions.
• Broker data: Quality can vary, data might have errors, duplicates, gaps.

Analysis

• Various tools: Custom programs, Python, R, MATLAB, commercial platforms, and APIs.
• Analysis needs to be robust; free data often unreliable.

Trading Execution

• Different approaches: build your own platform or use existing ones.
• Benefits of using integrated platforms: Simplifies data acquisition, analysis, and execution.
• Popular Forex Platforms: MetaTrader, NinjaTrader, TradeStation, MultiCharts.

Popular Forex Platforms

MetaTrader (MT4 and MT5)

• The most popular for retail Forex trading.
• Languages: MQL4 (C-like scripting language).
• Issues: Poor data quality, not sophisticated, not suitable for multiple simultaneous orders.

NinjaTrader

• Popular for futures, growing in Forex segment.
• Languages: C#, .NET 3.5.
• Requires separate broker API integration; steep learning curve.
• Excellent analysis features but requires effort to set up.

TradeStation

• Popular for equities; uses EasyLanguage (simple, aimed at non-programmers).
• Integration and trade execution seamless but limited to one broker.

MultiCharts

• Clone of TradeStation; compatible but offers more features and allows selecting data feeds and brokers.

Other Broker-Specific Platforms

• Thinkorswim: Strong in equities and options, limited language capabilities.
• Custom programming often required for complex trading strategies.

Execution and Risk with Brokers

Common Issues

• Poor customer service and execution: Re-quotes, slippage, bad latency.
• Inconsistent platform reliability: Software crashes, bad banking relationships.

Broker Profit Models

• Commissions or trading risk: Majority use marked-up spreads.
• Retail traders generally lose more during high volatility, contributing to broker profits.

Algorithmic Trading Strategies

Basic Strategies and Tools

• Sentiment analysis: Valuable for predicting market moves (MyFXBook, Oanda positioning data).
• Simple trading rules and backtesting: Moving averages cross-over strategies as examples.
• Data quality critical: Good data sources essential for accurate backtesting and live trading.

Special Considerations for Currencies

Forex Market

• Highly liquid but competitive with institutional players.
• Influenced by macroeconomic factors, global news, interest rates.

Futures Market

• Commitment of Traders (COT) Report: Commercial vs. retail trader positions.
• Strategies built around long-term data, often involve portfolio management.

Exotic and Third World Currencies

• Generally avoided due to lack of liquidity and higher trading costs (wider spreads).
• Interesting but risky; better for special, large-scale strategies.

Programming and Platform Considerations

• Strong foundation in specific languages (e.g., Python, C#) beneficial.
• Platforms like NinjaTrader offer powerful analysis options and multi-broker support but come with a learning curve.
• Integration between platforms can be complicated and often requires tailor-made solutions.

Conclusion

• Understanding of platforms and market dynamics crucial for success in algorithmic trading.
• Balancing simplicity with analytical power is vital; choose tools that fit your trading style and expertise.