Please enable JavaScript.

Coggle requires JavaScript to display documents.

Algorithmic Trading (Overview (Concepts (Strategies (Arbitrage…

- - - - Use multiple strategies, each strategy will have a function corresponding to it
        
        only buy/sell shares when all strategy functions return true
      - Keep algorithm running on the most stable/ highest performing servers (AWS)
      - Start with trading only 1 stock
      - to prevent having to re-write a lot for the code, use a FP design pattern and and HTML form (or something like that)
    - - Udemy
        
        Python for Financial Analysis and Algorithmic Trading https://www.udemy.com/course/python-for-finance-and-trading-algorithms/
        
        covers the following:
        
        trading concepts
        
        programming & data science
        
        develop your own algorithms
        
        this course uses quantopian
        
        recommended by industry leaders
        
        Pre-requisites
        
        Complete Python Developer in 2020 https://www.udemy.com/course/complete-python-developer-zero-to-mastery/
        
        covers design patterns
        
        might have enough machine learning concepts for us to be able to introduce in the algorithm
        
        (assuming we want the algorithm to learn from past decision)
        
        don't have to complete the entire course (not worth it imo)
        
        My notes from this course:
        https://coggle.it/diagram/Xp9jEepGhy5oq9By/t/python-zero-to-mastery/e58ca6b81cdaf56a00fea61ccf8406c83c8b1d84f26989fe958809f4a977477c
      - Quantopian https://www.quantopian.com/home
        
        we can use this site to test our algorithm on live data
    - - Tradestation
      - Multicharts
      - NinjaTrader
  - - - Define a set of instructions to follow before placing a trade
        
        based on timing, price, quantity or mathematical model
    - - Buy 50 shares of a stock when it's 50 day moving average goes above the 200 day moving average
      - Sell shares when it's 50 day moving average goes below the 200 day moving average
    - - trades are timed correctly
      - can be used to backtest historical data
      - rules out human emotion
    - - long-term investors
      - short-term traders
      - systematic traders
        
        for ex. hedge funds
    - - Trend-following
      - Arbitrage Opportunities
        
        Joshua's gap strategy uses this
        
        Resource: https://www.investopedia.com/ask/answers/what-is-arbitrage/
      - Index Fund Rebalancing
      - Mathematical Model-based
      - Trading Range
      - Volume-weighted Average Price
      - Time Weighted Average Price
      - Percentage of Volume
      - Implementation Shortfall
      - Beyond the Usual Trading Algorithms
    - - computer programming
        
        more in Technical Skills section
      - network connectivity
      - access to market data
      - backtesting
        
        VERY IMPORTANT: test the algorithm on historical data to see how it would've performed
  - - - what's a stop order?
      - what's a limit order?
      - margin requirements for the market you want to trade in?
      - is the exchange regulated?
    - - paper trade is VERY VERY useful
    - - financial calculations
      - basic statistics
      - computing trading performance metrics
    - - exploit weaknesses by tricking the software
      - design, develop, test algo strategies
    - - most trading systems lose money in the long run, the ability to quickly change an algorithm is important
  - - - high frequency traders don't stand a chance against big companies