Part 18 is about using synthetic data to test models and build resiliency
I agree. But the risk of using synthetic data is that the real stock is stochastic in nature and you'll tend to overfit if the data isn't not generated through sufficiently sophisticated means.
I also tried creating a new SPY/VIX long or cash model trained on some synthetic data to see what I get: https://kniyer.substack.com/p/a-new-spyvix-long-or-cash-model-powered
I agree. But the risk of using synthetic data is that the real stock is stochastic in nature and you'll tend to overfit if the data isn't not generated through sufficiently sophisticated means.
I also tried creating a new SPY/VIX long or cash model trained on some synthetic data to see what I get: https://kniyer.substack.com/p/a-new-spyvix-long-or-cash-model-powered