Bitcoin SF Devs Seminar: Bitcoin Algorithmic trading with CoinTrader and R

2 Bitcoin market microstructure presentations by SF Bitcoin Dev members Tim Olson and Tom Johnson

Coin Trader

CoinTrader is a new open-source platform for algorithmically trading cryptocurrencies. Java-based Coin Trader is designed to explicitly deal with cryptocurrency-specific issues like fractional volumes, long trade delays, nonstandard fee structures, and float roundoffs, and it can already connect to many exchanges, collect and persist market data, export CSVs, and perform ad-hoc data queries. Accounting and order execution will soon be developed, along with basic arbitrage strategies, trading circuit breakers, and a backtesting harness. Tim Olson will give an architectural overview of Coin Trader and discuss some basic operational issues around algorithmic trading. https://github.com/timolson/cointrader

Tom Johnson

By simultaneously buying and selling irregularities among crypto and fiat currency prices, arbitrage helps smooth the volatility that impedes adoption of BTC as a medium of exchange. Tom will show the challenges, plus methods and code, for spotting and capitalizing on these trading opportunities. Two types of arbitrage: Single - (home/crypto currency pairs) and Triangular - (home/2 cryptos or home/crypto/fiat currency triads). A demo using R in AWS cloud with an API of live synthetic prices will illustrate this. And, Tom will show how using a CSV file from the trading platform, more advanced algorithms get created which can then generate buy or sell signals within the trading engine.