When presented with volatile markets, this volatility is also an amazing opportunity. Extracting VRP can be used as an additional play in your playbook. If you are into options and not familiar with Variance Risk Premium you can learn more here: Extracting Variance Risk Premium. Not only is it conceptually a simple concept to understand and implement, it is very profitable. You can follow along here VRP 350 AI as we are completely transparent on the current performance and what our members are experiencing. Reach out with any questions. We have Options Trading experts to help with any questions!
96% return every seven months. You have found the Holy Grail. Shouldn't wall street firms be paying you millions for your secret, wizardry, knowledge -- instead of selling this formula to failed trader pikers on a forum,
Looks like it basically boils down to betting on direction. Seems like you're ultimately trying to predict the short-term direction of the S&P500 using fairly common machine learning algorithms.
Yeah we display our recent trades you would receive by signing up. It is free for 2 weeks if your curious on the signals. We want to be very clear about the system and we describe in detail how it works here: VRP 350 Intro We hope you'll check us out and experience what our members do. BTW, Those last 10 trades have a profit in them as well.
We are traders and developers, helping traders. We actually don't believe in the Holy Grail. See more on that here: Statistical Arbitrage Do check us out, and let us know if you have any questions. Thanks!
The architecture of your ML pipeline looks interesting. Curious what you used to implement it. I'm just getting started with the PyTorch and FastAI libraries. What library do you use to implement RL? Still curious why you are using a directional option strategy instead of delta neutral if you're betting that options (implied volatility) are overpriced relative to realized volatility. Is the logic that the eventual drop in IV will cause a rise in the underlying asset?
In terms of implementation, we used Keras/Tensorflow (Keras is really simple to set up and get up and running fast). Despite appearances, the strategy is not truly directional. The underlying neural net is in fact computing a probabilistic distribution function, and the only thing we are doing is pricing some replicated binaries across the curve and deciding if they are overpriced or not.
Why is something this sophisticated not being marketed institutionally? first New York would fund you 10x more than the entire elitetrader population as AUM.