Hi Quiet1, can you please describe the serious risk I have taken? The only risk I can see that I've taken is potentially being overexposed to VX. Do you think that I am dangerously exposed to VX? Thanks.
Fully automated futures trading
- Thread starter globalarbtrader
- Start date
5 Vix contracts is $5000 per VIX point and about $55000 in initial margin currently. If VIX futures had a small crash to say 23 you would be seriously down on your account, possibly even automatically stopped out by your broker due to an increase in margin requirement.
And that's just a bad outcome not the worst case scenario. You can't look at VIX as a normal futures market. It does not matter that the biggest 1 day rise in history was X%, you should be comfortable with VIX rising maybe 50 to 100% more than it ever has before. Basically you need to ensure that it's virtually impossible for the market to wipe you out. As it is, it's quite easy to imagine fairly plausible scenarios where your account loses 50%++ by large but not catastrophic geopolitical events.
And that's just a bad outcome not the worst case scenario. You can't look at VIX as a normal futures market. It does not matter that the biggest 1 day rise in history was X%, you should be comfortable with VIX rising maybe 50 to 100% more than it ever has before. Basically you need to ensure that it's virtually impossible for the market to wipe you out. As it is, it's quite easy to imagine fairly plausible scenarios where your account loses 50%++ by large but not catastrophic geopolitical events.
Thanks for that explanation, Quiet1. Based on what you say, I fear that I may have some error in my implementation of the system inspired by GAT's system, given that it has called for me holding 5 short VX on a $130k account at 30% volatility. I think I may have to liquidate, and revert to paper trading the system, until I see where my error is. Thanks for the warning.
It sounds like you have not even done a proper backtest yet. Do not jump straight into the trading a system live until you have tested all your rule implementation against historical data. A bug in your code can wipe out your account
I hope I'm not hijacking this thread. If so, please let me know.
To summarize, I am using a lot of GAT's excellent open source code to run my system. I am using 8 instruments, with $130K, running at 30% volatility: Vix, Nasdaq, Leanhogs, GBP, Platinum, US2, Eurodollar, GAS_US.
Here is the backtest with a sharpe of .82, and some other statistics. The annual returns assume a constant capital of $130K.
Sharpe: 0.8237007586663045
[[('min', '-35.16'), ('max', '38.82'), ('median', '-0.0001634'), ('mean', '0.1124'), ('std', '2.182'), ('skew', '-0.4452'), ('ann_mean', '28.76'), ('ann_std', '34.92'), ('sharpe', '0.8237'), ('sortino', '1.04'), ('avg_drawdown', '-19.87'), ('time_in_drawdown', '0.9454'), ('calmar', '0.3123'), ('avg_return_to_drawdown', '1.448'), ('avg_loss', '-1.256'), ('avg_gain', '1.524'), ('gaintolossratio', '1.213'), ('profitfactor', '1.176'), ('hitrate', '0.4922'), ('t_stat', '5.421'), ('p_value', '6.044e-08')], ('You can also plot / print:', ['rolling_ann_std', 'drawdown', 'curve', 'percent', 'cumulative'])]
1975-12-31 16584.943069
1976-12-31 57007.095302
1977-12-31 185288.524317
1978-12-31 9038.766412
1979-12-31 28444.836655
1980-12-31 -3524.120744
1981-12-31 90194.841088
1982-12-31 52757.341149
1983-12-31 1222.886848
1984-12-31 49404.534208
1985-12-31 43364.894901
1986-12-31 59814.564226
1987-12-31 57271.110810
1988-12-31 -34722.603749
1989-12-31 -6298.312574
1990-12-31 100327.008972
1991-12-31 96966.323021
1992-12-31 -1769.737457
1993-12-31 70971.363596
1994-12-31 14731.283840
1995-12-31 10882.497068
1996-12-31 81813.083127
1997-12-31 -21759.192076
1998-12-31 62131.275456
1999-12-31 -54857.617955
2000-12-31 88060.679064
2001-12-31 10110.929835
2002-12-31 74961.156730
2003-12-31 50043.099646
2004-12-31 61388.005078
2005-12-31 -30585.911546
2006-12-31 -32560.409092
2007-12-31 -40131.714264
2008-12-31 65827.532510
2009-12-31 22804.673778
2010-12-31 138670.035128
2011-12-31 67260.134254
2012-12-31 33172.274671
2013-12-31 -811.251820
2014-12-31 72000.055353
2015-12-31 28589.906901
2016-12-31 3245.530233
2017-12-31 27333.300763
Would be interested in your thoughts. Thanks.
To summarize, I am using a lot of GAT's excellent open source code to run my system. I am using 8 instruments, with $130K, running at 30% volatility: Vix, Nasdaq, Leanhogs, GBP, Platinum, US2, Eurodollar, GAS_US.
Here is the backtest with a sharpe of .82, and some other statistics. The annual returns assume a constant capital of $130K.
Sharpe: 0.8237007586663045
[[('min', '-35.16'), ('max', '38.82'), ('median', '-0.0001634'), ('mean', '0.1124'), ('std', '2.182'), ('skew', '-0.4452'), ('ann_mean', '28.76'), ('ann_std', '34.92'), ('sharpe', '0.8237'), ('sortino', '1.04'), ('avg_drawdown', '-19.87'), ('time_in_drawdown', '0.9454'), ('calmar', '0.3123'), ('avg_return_to_drawdown', '1.448'), ('avg_loss', '-1.256'), ('avg_gain', '1.524'), ('gaintolossratio', '1.213'), ('profitfactor', '1.176'), ('hitrate', '0.4922'), ('t_stat', '5.421'), ('p_value', '6.044e-08')], ('You can also plot / print:', ['rolling_ann_std', 'drawdown', 'curve', 'percent', 'cumulative'])]
1975-12-31 16584.943069
1976-12-31 57007.095302
1977-12-31 185288.524317
1978-12-31 9038.766412
1979-12-31 28444.836655
1980-12-31 -3524.120744
1981-12-31 90194.841088
1982-12-31 52757.341149
1983-12-31 1222.886848
1984-12-31 49404.534208
1985-12-31 43364.894901
1986-12-31 59814.564226
1987-12-31 57271.110810
1988-12-31 -34722.603749
1989-12-31 -6298.312574
1990-12-31 100327.008972
1991-12-31 96966.323021
1992-12-31 -1769.737457
1993-12-31 70971.363596
1994-12-31 14731.283840
1995-12-31 10882.497068
1996-12-31 81813.083127
1997-12-31 -21759.192076
1998-12-31 62131.275456
1999-12-31 -54857.617955
2000-12-31 88060.679064
2001-12-31 10110.929835
2002-12-31 74961.156730
2003-12-31 50043.099646
2004-12-31 61388.005078
2005-12-31 -30585.911546
2006-12-31 -32560.409092
2007-12-31 -40131.714264
2008-12-31 65827.532510
2009-12-31 22804.673778
2010-12-31 138670.035128
2011-12-31 67260.134254
2012-12-31 33172.274671
2013-12-31 -811.251820
2014-12-31 72000.055353
2015-12-31 28589.906901
2016-12-31 3245.530233
2017-12-31 27333.300763
Would be interested in your thoughts. Thanks.
This is exactly what I meant with one of my earlier replies to you.
I'll have to take a look at how to derive max historical drawdown.
I am willing to accept the risk of that 77% drawdown that you noted, given the potential upside of +20% annual returns. I have my other capital in etfs, and all of this is meant for retirement which is maybe 30 years away.
Are you prepared to top up your capital for 3 consecutive years (2005-2007)? As per your earlier post, you would have to increase your capital to $130k at the start of each year.