Fully automated futures trading

7% down here.
I'll say this is karma for me bragging about missing that carnage at the end of November. :D

Interesting thing is that 3/4 of the loss are from single long heating oil contract. Yep, I'm that unlucky to be long still.
Fun fact - I introduced DO on Monday. My system before that got me out of HO on Friday due to the increased vol, but DO said go long, so I let the system go. I'm thinking of asking for my money back, but my supplier of DO algorithms is a retired solo bloke just outside of London, pretty sure I'm not getting anything :D
That's ok though, I know these days happen, yesterday was 4.6% up day, that makes this easier to stomach. Still up about 8.5% YTD.

I am wondering why though, why this reversal. I haven't seen anything in the news to warrant this.
 
wopr said:
7% down here.
I'll say this is karma for me bragging about missing that carnage at the end of November. :D

Interesting thing is that 3/4 of the loss are from single long heating oil contract. Yep, I'm that unlucky to be long still.
Fun fact - I introduced DO on Monday. My system before that got me out of HO on Friday due to the increased vol, but DO said go long, so I let the system go. I'm thinking of asking for my money back, but my supplier of DO algorithms is a retired solo bloke just outside of London, pretty sure I'm not getting anything :D
That's ok though, I know these days happen, yesterday was 4.6% up day, that makes this easier to stomach. Still up about 8.5% YTD.

I am wondering why though, why this reversal. I haven't seen anything in the news to warrant this.
More...

You ain't blaming the algo buddy, it's got an alibi.

DO sold me out of heating oil on Friday, again due to increased vol.

Code: 
         instrument_code          strategy_name    trade parent_reference_price parent_limit_price mid_price side_price offside_price limit_price filled_price        delay bid_ask    execution versus_limit versus_parent_limit total_trading
order_id                                                                                                                                                                                                                                       
32003            HEATOIL       dynamic_TF_carry     [-1]                  3.182               None    3.2256     3.2249        3.2263        None       3.2249       0.0436 -0.0007           -0          NaN                 NaN       -0.0007

I note in passing that Heating oil is still higher than that price, so maybe not the best decision...

In all seriousness, I obviously still had other energy risk (mini WTI and last day Brent) on which got me in a fair amount of hot water, although not as much. I'm guessing that your system didn't see a better way of maintaining the energy long. Path dependence, slightly different signals or instruments, maybe different capital.... could be anything.

If it was a small capital / big contract size problem then I wonder if the risk limit I use to limit notional exposure on a single contract (described here https://qoppac.blogspot.com/2022/02/exogenous-risk-overlay-take-two.html) would have helped (see from 'Note that it's still possible for instruments to exceed the leverage limit, but this is dealt with by the off line hard position limits' onwards).

A piece of advice; once you trust your new system, stop monitoring your old one. It's a waste of time and regretful tears.

Rob
 
globalarbtrader said:
If it was a small capital / big contract size problem then I wonder if the risk limit I use to limit notional exposure on a single contract (described here https://qoppac.blogspot.com/2022/02/exogenous-risk-overlay-take-two.html) would have helped (see from 'Note that it's still possible for instruments to exceed the leverage limit, but this is dealt with by the off line hard position limits' onwards).
More...
This is next up on my list!
I'm obviously joking, I'm pretty happy with the DO stuff, backtests are looking great and I should give it more than 2 days to make judgements :)

I do have an interesting implementation that I wanted to write about. I don't run pysystemtrade, but I "hook into" it for some behavior, DO computation being one. Having done this, I think now that the easier path would have been to just implement a pysystemtrade data source for my data store, but I wanted to fully understand it so I use everything from my system, compute all the inputs, and call into objectiveFunctionForGreedy for actual optimization. I also run Python 3.10, so don't ask me how I'm plugging that in :D
 
globalarbtrader said:
Here are the returns again since the google sheets work is borked for some reason

Can't believe I've been doing this for over 2000 business days; coming up to 8 years!

Code: 
2014-04-08  
2014-04-09  
2014-04-10  
2014-04-11    -0.95%
2014-04-14    -0.36%
2014-04-15    1.13%
...
More...

Hi GAT,

Are the daily figures in the above post actual account figures, or a back test or a combination of both?
 
globalarbtrader said:
If it was a small capital / big contract size problem then I wonder if the risk limit I use to limit notional exposure on a single contract (described here https://qoppac.blogspot.com/2022/02/exogenous-risk-overlay-take-two.html) would have helped (see from 'Note that it's still possible for instruments to exceed the leverage limit, but this is dealt with by the off line hard position limits' onwards).
Rob
More...

I think this would've saved me yesterday, going to experiment with something similar. Although it probably would've prevent the initial opening of my Silver long which went completely in my face. I realised I had too risk on by holding the contract when the account madly increased on Tuesday, but if it's going in your favour it's hard to switch it off, then of course it turned around and I gave it all back. It was a bitter slice of humble pie.
 
You must log in or register to reply here.
Back
Top