Hi! Never mind the title. I know this is not algo trading! As a hobbyist/enthusiast in stock market movements, I was recently curious how a simple MA crossover strategy would perform as compared to a random decision based strategy. As an example I implemented a basic stream processing application. This is more of an exercise in software development principles than trading.
**WARNING: Please do not apply this with real money (or do so at your own risk!)
We read data as events from a binance provided websocket aggTrade streams. Two wallets are initialized (with virtual money) and each is subjected to a different strategy. One is a simple moving averages based crossover strat that compares two moving averages (long and short) which are evaluated on each event and makes a BUY/SELL decision upon crossover (More here..) and decides to HOLD otherwise. The other is a random decision strategy which randomly chooses either of BUY/HOLD/SELL for each input event, agnostic of event parameters.
The strategies are also agnostic of any wallet and make their decisions based on aggregated trade signals only. An executor connects a strategy to a wallet and executes an actual transaction on the wallet depending on the strategy output. For simplification, a fixed amount executor is implemented that decides to buy a fixed amount of an asset corresponding to a BUY/SELL signal. As output, we display the wallet balances (both asset holding and money) which are refreshed with each event
Just running go run . should do it! :)