Quantopian's community platform is shutting down. Please read this post for more information and download your code.
Back to Community
Best custom benchmark to use for market-neutral strategies?

I've been using Quantopian this year for an independent project as part of my MBA. I've been modelling academic papers and trying to implement them as small-scale versions of hedge fund strategies.

One problem I have is that, for a market neutral strategies (for simplicity let's say that is beta -0.30 to 0.30, in line with the Quantopian Open guidance), SPY is the wrong benchmark. SPY will typically beat an unlevered market neutral strategy, even if it produces moderate alpha, because SPY is taking more market risk (it has a beta of 1, of course). My supervisor said typically you'd compare a back test on a 'market neutral' strategy to the HFRI or HFRX (hedge fund) index. There isn't a robust ETF to track those indicies (I think).

Does anyone have ideas for things we can use as a benchmark that are compatible with Quantopians set_benchmark function (i.e. are a traded stock)?

1 response

If the return has no beta, ten its alpha. The proper benchmark for a true market neutral portfolio is the risk-free rate. 30-day t-bills would work.