Fully automated futures trading

Flash figure for 2019/20 tax year finishing yesterday: +35.7%

Attribution: Futures trading +39.7%
Stocks plus hedge: -4.0%

I'll do the more exhaustive analysis on my blog in the near future, which will include the total shitshow that is my long only investment portfolio. But this is the best year I've had since 2014/15.

I'm also now on the 'front page' of the fundseeder leaderboard, in 10th place. The highest I've been since I switched the asset base calculation to a more realistic method that unfortunately penalised me quite badly.

GAT
 
globalarbtrader said:
Flash figure for 2019/20 tax year finishing yesterday: +35.7%

Attribution: Futures trading +39.7%
Stocks plus hedge: -4.0%

I'll do the more exhaustive analysis on my blog in the near future, which will include the total shitshow that is my long only investment portfolio. But this is the best year I've had since 2014/15.

I'm also now on the 'front page' of the fundseeder leaderboard, in 10th place. The highest I've been since I switched the asset base calculation to a more realistic method that unfortunately penalised me quite badly.

GAT
More...
Congrats. That’s impressive to say the least. After a good start middle of last year, my system unlike GAT’s (and the other mini-GATs on this forum :)) has struggled during the exceptional volatility during the last few weeks. This I discovered recently was largely due to badly specified trading buffers I was using in my system which prevented my system exiting loss making positions sufficiently quickly (essentially systematically replicating at least in part the human bias of holding on to loss making positions). I also did not benefit from the move in markets like hogs as I don’t trade meats for personal reasons. I am therefore only marginally up on my futures trading since I started (I would be up a lot more if had had smaller buffers). This is now corrected. On a more positive note my discretionary side bets (short s&p, long gold) on this occasion have paid off so far, so my overall returns haven’t been too bad (+10 pct since jun 19) but my discretionary track record is patchy at best so I am not banking on being able to repeat that.
 
Hello Rob
I am reading your book Leveraged Trading that has a lot of new concept for me that I try to understand.
Currently I am bit stuck with “Table 11: Chances of a given annual loss when running the Starter System”. How do you come to this table from standard deviation and mean?

I would assume you would use annualized standard deviation and annualized mean.
For example annual mean 5 % and annualized standard deviation 12 % that should be closed to the results of Starter system. To get the probability of 20 % loss I would use Excel formula =NORMDIST(-20,5,12,TRUE)*100, but this give me quite unrealistic result of 1,87%. In your book you have more realistic 5,8%.

Can you please give me hint what I am doing wrong?
Thank you
 
quiper said:
Hello Rob
I am reading your book Leveraged Trading that has a lot of new concept for me that I try to understand.
Currently I am bit stuck with “Table 11: Chances of a given annual loss when running the Starter System”. How do you come to this table from standard deviation and mean?

I would assume you would use annualized standard deviation and annualized mean.
For example annual mean 5 % and annualized standard deviation 12 % that should be closed to the results of Starter system. To get the probability of 20 % loss I would use Excel formula =NORMDIST(-20,5,12,TRUE)*100, but this give me quite unrealistic result of 1,87%. In your book you have more realistic 5,8%.

Can you please give me hint what I am doing wrong?
Thank you
More...

A mean of 5% is too high, as that gives a Sharpe Ratio of 0.42.

Assumptions are SR 0.24, standard deviation 12%, then your expected return will be 0.24*12% = 2.88% per year

Using a simple excel calculator I make that a 2.8% chance of a 20% loss.

However the actual results in the book are bootstrapped from backtested returns, and so account for jumps and non Guassian returns.

GAT
 
Thank you. Is it correct to assume that different trading systems with the same risk target will have similar probabilities like the one published in Table 11 of the book? Or do you think different trading styles will lead to different probabilities with the same risk target?
 
quiper said:
Thank you. Is it correct to assume that different trading systems with the same risk target will have similar probabilities like the one published in Table 11 of the book? Or do you think different trading styles will lead to different probabilities with the same risk target?
More...

Will depend on the system, but the table 11 values are fairly conservative and not a bad guide.

GAT
 
Going back to the question of bargain-hunting, is it normal that when I
globalarbtrader said:
I keep this portfolio pretty simple. 30% stop loss. My UK universe is FTSE 350. The filter is as desribed in Smart Portfolios.

Forward PE < 12
EPS growth > 0
Forecast yield > 4%
PTBV < 1.5
Forecast div. Cover > 1.2
Gearing < 50%


I don't do MVO, just buy equal cash weight.

GAT

PS Looks like I was a bit early calling the bottom :)
More...

Going back to the question of bargain-hunting, is it normal that when I'm setting all the scanner parameters as described almost all companies I'm getting are financial (banks and insurances), only 4 out of 32 are in some other business. Is it really that all the "best to buy" companies in the US at the moment are financial or these parameters are naturally skewing selection to finance sector?
Or I messed up some of the parameters of the scanner ?
I did set in addition Market Cap. over 2bln, though..

https://finviz.com/screener.ashx?v=...pe_u10,fa_ltdebteq_u0.5,fa_pb_u2,geo_usa&ft=4
 
Last edited:
Kernfusion said:
Going back to the question of bargain-hunting, is it normal that when I'm setting all the scanner parameters as described almost all companies I'm getting are financial (banks and insurances), only 4 out of 32 are in some other business. Is it really that all the "best to buy" companies in the US at the moment are financial or these parameters are naturally skewing selection to finance sector?
Or I messed up some of the parameters of the scanner ?
I did set in addition Market Cap. over 2bln, though..

https://finviz.com/screener.ashx?v=...pe_u10,fa_ltdebteq_u0.5,fa_pb_u2,geo_usa&ft=4
More...
It could be that many of the companies you would be interested in have meanwhile decided to not pay out dividends. Those companies prefer to keep the cash in their pocket, in order to be better able to weather the ongoing storm. While this is prudent behavior, in my view, it would exclude these companies from your search as you require a dividend yield of at least 4%.
What I'm trying to say is that filter criteria which are suitable in normal times are probably not suitable during a special situation as the current corona virus crisis and the economic implications on companies worldwide.
 
Last edited:
Rob's post (https://elitetrader.com/et/threads/fully-automated-futures-trading.289589/page-205#post-5038211) and Kernfusion's question the other day pushed me to think about portfolio margin accounts, so wanted to ask a few questions here to make sure I understand how that works.
Currently, my system is out of most positions, and margin is about 10% right now, so I have a lot of cash sitting around. I was planning on switching to portfolio margin, so I can do something similar to what Rob did, buy some short term gov bonds to get some yield. I could do that now too, but much less, as that position would eat up a good chunk of the margin, and also, I have to keep in mind that futures margin will go up again once the system opens positions.

Are there any downsides to doing that? What am I missing?
My thinking is that as long as I don't use the additional "buying power" to increase leverage and get into bigger positions by increasing account risk, that should be ok.
If it's relevant, I'm using IB.
 
wopr said:
Rob's post (https://elitetrader.com/et/threads/fully-automated-futures-trading.289589/page-205#post-5038211) and Kernfusion's question the other day pushed me to think about portfolio margin accounts, so wanted to ask a few questions here to make sure I understand how that works.
Currently, my system is out of most positions, and margin is about 10% right now, so I have a lot of cash sitting around. I was planning on switching to portfolio margin, so I can do something similar to what Rob did, buy some short term gov bonds to get some yield. I could do that now too, but much less, as that position would eat up a good chunk of the margin, and also, I have to keep in mind that futures margin will go up again once the system opens positions.

Are there any downsides to doing that? What am I missing?
My thinking is that as long as I don't use the additional "buying power" to increase leverage and get into bigger positions by increasing account risk, that should be ok.
If it's relevant, I'm using IB.
More...
You mention "portfolio margin" and I assume you compare it to Reg-T margin at IB. My account is Reg-T and there was a what-if tool to see how it would compare to portfolio margin. To my surprise was portfolio margin lower than Reg-T margin, so I decided to not change over.
Even with portfolio margin can you use a large portion of your cash for stock purchases. As long as your futures positions remain small will you not hit the maximum. And when you do start to hit the maximum you could decide to reduce those stock positions. The current interest rates for excess cash is rather low, so it should be possible to get a bit better return by using some stock investments.
 
You must log in or register to reply here.
Back
Top