Fully automated futures trading

I can plug it for you, just yesterday I referenced the tactics section. I've been trying to adjust how I handle STIRs and that section is useful. In fact I'd say there's loads of information in there even fairly experienced futures traders did not know.

 

Thank you, that's kind

Rob

 

Can't say for everyone else, but I sure think it could be fun

 

True, I never thought of it this way (btw, I am @sle - just this is my new handle as I recycled the old one).

On a different note, was wondering - would it be advantageous to use implied vol x historical correlation matrix as opposed to historical covariances to arrive at sizing in your model or did you find/think that it does not make much of a difference?

 

It would be an astonishing amount of work, and isn't even possible in many cases where there aren't liquid options markets.

A fun thing to do is plug a forward looking forecast of volatility into your backtest, and see what happens to performance. Skew will improve, but Sharpe ratio probably won't. Unless you're running at crazy leverage you won't get much benefit.

Rob

 

Of the top of my head, I can't think of a liquid futures market (at least in the US) that does not have a liquid(ish) options market. My prior is that both sharpe and sortino will improve somewhat. Though it's hard to say by how much - an approach like that will reduce exposure to some of the volatile reversals/gaps (especially in commodities) but will also avoid allocating to some trends. So I can see if it's a lot of workand also the length of backtest will reduce significantly, it's not worth it.

 

From where would you get historical implied vol data?

 

Me personally? I'd get it from from our futures vol database or, if I did not have that, from Bloomberg (they have constant maturity series for pretty much every futures options). If I was a retail trader, I'd probably map the futures onto optionable ETFs/indices and use something like ORATS.

 

Any obvious downsides to create a ladder along the term structure? I do this (for instruments that have several liquid contracts) when holding multiple contracts in an effort to reduce how often I need to roll and thus trading costs.

 

I often hold two contracts to smooth rolling (see AFTS 'tactics' chapters), but any more would make things too complicated from an infrastructure perspective.

Rob