ES Trading Performance

[QuantWizard]

Hi all,

I have been trading ES with an automatic trading system for the last year, please have a look and let me hear your verdict!

The attached image shows actual net performance as from broker.

For the actual trading, performance stats are as follows:
KPI Actual / Benchmark
===================== ========== ===========
Capital invested: $1370.00 / $1370.00
Current equity: $5269.79 / $1594.06
Net ROI 1 day (+/-): $183.48 / $16.75
Net ROI 1 day (%): 3.62% / 1.06%
Net ROI 1 week (+/-): $218.28 / $3.75
Net ROI 1 week (%): 4.34% / 0.24%
Net ROI 1 month (+/-): $1854.84 / $34.25
Net ROI 1 month (%): 54.64% / 2.17%
Net ROI total (+/-): $3899.79 / $224.06
Net ROI total (%): 284.66% / 16.35%
Ann. return (%): 272.79% / 15.95%
Ann. volatility (%): 198.14% / 12.79%
Modigliani measure: 3.23% / 2.99%
Alpha: 240.41% / 0.0%
Ann. Sharpe Ratio: 1.53 / 1.23
Max drawdown: 68.67% / 8.87%
Win-to-loss ratio: 1.01 / N/A
Beta: 2.89 / 1.0
R Squared: 0.37 / 1.0
VaR (5%, 1 Month): 90.65% / 5.92%

 

[gmst]

You are not losing money while you are paying comm and slippage. So, you are in top 20% of traders. To become consistent you need to do lots more work. Lets see what others tell you.

 

[QuantWizard]

Quote from gmst:

You are not losing money while you are paying comm and slippage. So, you are in top 20% of traders. To become consistent you need to do lots more work. Lets see what others tell you.

Thanks for your feedback. After the big drawdown you can see in Aug, I adjusted a few parameters to make it less aggressive, and ever since kept things more or less constant.

The interesting thing is, if I would have stayed out of the market on days where the VIX is 16 or higher, the drawdown would have been avoided all together. However, by doing so, I'd be out of the market one-third of the time which feels a bit boring (VIX has been below 16 from Aug more or less all the time until now).

Will keep you updated.

 

[gmst]

Quote from QuantWizard:

Thanks for your feedback. After the big drawdown you can see in Aug, I adjusted a few parameters to make it less aggressive, and ever since kept things more or less constant.

The interesting thing is, if I would have stayed out of the market on days where the VIX is 16 or higher, the drawdown would have been avoided all together. However, by doing so, I'd be out of the market one-third of the time which feels a bit boring (VIX has been below 16 from Aug more or less all the time until now).

Will keep you updated.

obviously there are few things here going for you. First - you are sticking to your rules. Not many people can do this. This is in itself a big achievement.

Being bored is one of the hardest thing to control. I would suggest put all your energies towards diversifying into other products or discovering additional strategies. That is the best antidote to boredom, imo.

 

[QuantWizard]

Quote from gmst:

obviously there are few things here going for you. First - you are sticking to your rules. Not many people can do this. This is in itself a big achievement.

Being bored is one of the hardest thing to control. I would suggest put all your energies towards diversifying into other products or discovering additional strategies. That is the best antidote to boredom, imo.

Actually, I've experimented with other strategies as well, e.g. hidden markov models, factor models etc., but all of which has not proven to perform equally well in back tests.

I strongly believe that finding good strategies is just as much luck and hard work in addition to talent, so it's just a matter of time of capacity at the end of the day.

What I'm looking at right now is an "anti-fragile" approach à la Nassim Taleb - since I'm having 70% cash as a cushion, I'm thinking about a "safe way" of leveraging that part.

My most promising lead for the moment are high-yield bonds (held to maturity), where I'm currently simulating effects from default risks, so as to generate a yearly return of 8-10%; this should be the "realistic case" of a Monte Carlo simulation, and the strategy could be designed to render an acceptable "worst case" (e.g. one-in-a-million chance of 50% loss etc.).