Hi I am new here and I am trying to interpret the statistics that Quantconnect returns for my Crypto algorithm   backtests. I have algorithms that are returning strange statistic values for Alpha. i.e. 17.067 and CAGR i.e. 1068734551174.1%   I can go look into the  Quantconnect code but I thought I would ask before doing so in the hope of saving some time. 

How is Quantconnect calculating Alpha for the Cryptocurrency marketplace?

If we take the definiition of Alpha to be: “Alpha gauges the performance of an investment against a market index or benchmark which is considered to represent market movement as a whole. A positive alpha of 1.0 means the fund has outperformed its benchmark index by 1%.”   

What market index or benchmark is it using for Crypto?  Is it looking at the top 100 cryptocurrencies in coinmarketcap dot com? Is it just looking purely at some calculation based on BTCUSD price action?

Or is it the case that we are meant to create our own benchmark for it to compare against our algorithms?

I don’t believe CAGR can really be applied to the cryptocurrency market and I would prefer to use something based on market and risk like Net R Return or CAPM due to all the complex market forces at work. So for now I am just ignoring that nonsensical number it is returning for CAGR.  But if this number were to be believed I would be happy to be proven wrong. :-)

Thanks in advance for your assistance with this it is greatly appreciated.

Author