Back

How to tell if live/paper trading results match historical performance?

As the title says, I'm paper trading my strategy as a way to see if the performance in the backtest was over-fitted and not accurate to real life. However, I ran into a problem: How do I quantify the performance? My strategy could have a 5% drawdown and it could potentially be the day after I start forward testing it. How can I tell if this is normal drawdown or if it's abnormal so I can pull the plug?

I've seen graphics before where someone will show a strategy's equity line and then forecast it out into the future with some confidence intervals. So if the live trading results go outside these bounds indicating that the performance is not expected. How can I do this with Quantopian?

Update Backtest







0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Hi Lukel,

One way to classify a drawdown during forward testing as "abnormal" would be to add a MaximumDrawdownPercentPortfolio risk management model with its `maximumDrawdownPercent` parameter set to the backtest's maximum drawdown.

In regards to constructing a confidence interval for the equity curve, we can do this inside a custom risk management model. See the attached backtest for an example which creates this confidence interval based on the equity curve's standard deviation. There are many other ways of forecasting a confidence interval like this. A simple search of time series forecasting models yields several options worth trying.

Best,
Derek Melchin

0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Update Backtest





0

The material on this website is provided for informational purposes only and does not constitute an offer to sell, a solicitation to buy, or a recommendation or endorsement for any security or strategy, nor does it constitute an offer to provide investment advisory services by QuantConnect. In addition, the material offers no opinion with respect to the suitability of any security or specific investment. QuantConnect makes no guarantees as to the accuracy or completeness of the views expressed in the website. The views are subject to change, and may have become unreliable for various reasons, including changes in market conditions or economic circumstances. All investments involve risk, including loss of principal. You should consult with an investment professional before making any investment decisions.


Loading...

This discussion is closed