Jeff Holden develops algo trading models at SMB Capital in New York. He utilizes CloudQuant extensively during his research and development process. In this video he details his process for developing an idea into a model.

SMB Quant Video 1 – Developing a Model

SMB Capital recently started a new series of Videos/Podcasts where they interview Quants both inside and outside of their company.
In the first video, Jeff Holden of SMB steps through his process for developing a model :

  • Data Curation – What data do I need for this model?
  • Feature Analytics – Taking informative features and turning them into actual investment algorithms
  • Strategy – Take the analysis and turn it into a Theory – an understanding of WHY it works, what is the reason or motivation that causes the other trader on the opposite side to lose money to us. Often you finds something other than what you were looking for or even the exact opposite.
  • Backtesting – Assess – Test in unusual market environments – try to break it.
  • Deployment – Testing the model with live market data, Optimization, Scaling and Portfolio selection.
  • Q&A Session

Applications used : CloudQuant, Jupyter Notebooks, Gr8Trade (KTG Group), Gr8Py (KTG Group)

Duration : 51m