Fully automated futures trading

[djames]

IMHO with listed options you're not trying to do 'pricing model' arbitrage as you describe. Instead you're trying to better predict the path of realised and/or implied vol better than the other guy; either in an absolute sense or a relative sense.

GAT

thanks GAT, do you have any references to dive deeper?

 

[djames]

Just those two. I got around 0.7 as well, with equal weights in the portfolio. Then I bootstrapped the portfolio to find better weights, and that improved things. I'm a little concerned about overfit (I don't like this method), but I don't think equal weights is right either.

You can see my weights here:

Cheers, interesting. I'm bootstraping the weights also but I'm only trading 8 instruments at the moment so that likely explains the difference in Sharpe more than anything else.

 

[isotope1]

Cheers, interesting. I'm bootstraping the weights also but I'm only trading 8 instruments at the moment so that likely explains the difference in Sharpe more than anything else.

Definitely. What instruments are you trading?

 

[Elder]

I am surprised to see the high weighting in sugar given the returns data, but maybe I have bad data. Also presumably the sharpe comparison only makes sense if the two of you are trading a similar number of instruments?

Deleted my post as there seems to have been a lag at my end on seeing djames’ response (should have pressed refresh I guess)

 

[djames]

Definitely. What instruments are you trading?

corn, eurodollar, eurostox, gas, 3ktb, mxp, plat, v2tx

only been running a month(ish) most performance coming from eurodollar and v2tx

 

[djames]

thanks GAT, do you have any references to dive deeper?

Here's a reference which I found useful for numpty options traders: http://www.bespokeoptions.com/blog/2015/10/06/implied-volatility-with-c-and-python-pt-1

 

[HobbyTrading]

corn, eurodollar, eurostox, gas, 3ktb, mxp, plat, v2tx

only been running a month(ish) most performance coming from eurodollar and v2tx

Out of curiosity: did you get these instruments from @globalarbtrader 's blog (qoppac)? Or did you determine this set yourself?

 

[djames]

Out of curiosity: did you get these instruments from @globalarbtrader 's blog (qoppac)? Or did you determine this set yourself?

inspired by GATs blog, none of these are thresholded

 

[Napoleon@Dynamite]

I did some work on this. Two of the obvious options are:

• Mean reversion over shorter time scales. Trends tend to work on periods of more than a month, mean reversion in the shorter term.
• Trading a spread - e.g. long nasdaq short s&p.

I couldn't get either to work really: the short time scale stuff requires trading too fast (although if you set up your system to be realtime and used limit orders like a market maker, I expect it could work).

Trading the spread- I went off this because it looked like I'd have to have a huge position in order to make a small profit; something I'm not comfortable with.

Currently it looks like I'm anticipating a Sharpe of around 0.9. I don't know of any hedge funds that do consistently better, so I'm not sure what the benefit would be doing much more.

@isotope1, I have a backtested sharp of around .9+ using 15 instruments, and using GAT's 3 slowest emwas and carry with bootstrapping. I recently added GAT's breakout rule, and my sharpe dropped to .86. I wonder if that's a statistically significant difference, and if I should keep the rule or not.

 

[HobbyTrading]

inspired by GATs blog, none of these are thresholded

Be aware that those instruments were chosen based on value volatility of one contract of each at the time that blog post was written. Such volatility changes over time. For example: in the category energy had gas at that time a lower value volatility than crude oil. Nowadays would it be better to choose crude oil (CL) instead of natural gas (NG): CL has a lower value volatility per contract than NG. The same applies to forex contracts: MXP is currently not the contract with the lowest value volatility.