Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
20 Sharpe!! - Is that even possible?

Here is an algorithm on cryptocurrencies (long/short dollar neutral portfolio). I am using following commissions and slippage and rebalancing daily - see here

    context.set_commission(maker=0.001, taker=0.002)  
    context.set_slippage(spread=0.001)  

4 responses

Impressive. What pairs are you trading / which exchange and where do you get the data from? Also is it realistic given that it is very hard to short cryptos? (which assumptions do you use for repo rate?)

I am using enigma catalyst for backtests using data from poloniex with around 20 pairs. I haven't yet modeled the short cryptos and repo rate for them. Need to add that to the slippage I guess.

Pravin,
it is impressive. what factors did you use in selecting 20 pairs. most of the currencies have done really well in past 6 months. its difficult to back-test crypto as most of them are very new.

Thanks, Vivek. I cannot disclose details of the strategy but yes like you say backtesting cryptos is a challenge because of lack of data. Here is a longer backtest with lower turnover and longer holding horizon. It is based on bitcoin returns (instead of dollar returns). So effectively starting with a bitcoin neutral algorithm the returns are 250% in bitcoin terms after 2 years.