Hi everyone,
In this tutorial, we apply a gradient boosting model to predict intraday price moves when given a collection of technical indicators. The indicators we use are the RSI, MACD, and BB, but the algorithm is designed to allow for easy customization of these indicators. The backtest of the algorithm is shown below. For a thorough analysis of the strategy, check out the Strategy Library.
Best,
Derek Melchin
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.
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 Derek Melchin I read your post with interest as I am new to your platform and trying to implement and test the following similar approach for global equities either using a GBM or multi branch boosted tree. There is some evidence that multi branch trees beat binary trees and that these multi branch trees beat GBM on most metrics. That said testing this using GBM on QC woudl be a start.
Use classification instead of regression ie try to predict winners and losers by classifying the top say 30% of winners/outperformers as +1 and the bottom 30% losers/underperformers as -1. Almost all of the research and commentary I have seen use a regression approach and not classification.
As far as technical indicators go I woudl like to use 20 different unique technical indicators already available on the platform and normalize each indiactor value over its own lifetime history.
The target variable i.e. what I am trying to predict is the return of an equity over the folowing five trading days. Running this on a Friday after the close to predict the following Friday's return. This would amount to a weekly rebalance going long the top decile and short the bottom decile of equities. Execution ideally on the close but second choice woudl be VWAP.
Three sub models, all equally weighted woudl be:
a) short term model based on one month rolling lookback window
b) long term model based on one year/12 month rolling lookback window
c) seasonal model based on 10 year rolling lookback of the same calendar month.
Any guidance from your team appreciated as I am mid way through your boot camp but working on this as my first project on your platform.
Thank you
Hi Warren,
Welcome to QC!
When implementing this, utilizing the `Expiry.EndOfWeek` function will be helpful for creating insight durations. The source code is available here. To execute on the close, we need to create a custom execution model that creates MarketOnClose orders. For the 3 sub models above, I suggest taking advantage of our RollingWindow class.
I encourage continuing the bootcamp lessons. Then publish a backtest to the forum for assistance.
Best,
Derek Melchin
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.
Thank you Derek - thats helpful and will do
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.
To unlock posting to the community forums please complete at least 30% of Boot Camp.
You can
continue your Boot Camp training progress from the terminal. We
hope to see you in the community soon!
QuantConnect™ 2023. All Rights Reserved
|
Start New Discussion 
Back
Reward Discussion 
3
