https://preview.redd.it/qjd3lh3tkptg1.png?width=1637&format=png&auto=webp&s=469419e6750840200c634c26c3e10f6443a37140
https://preview.redd.it/3dk37kiukptg1.png?width=1637&format=png&auto=webp&s=834ae1b2e3d05a514ed4371b993c0bee70b8d273
I created a dashboard for analyzing my backtested 0DTE SPX trading [strategy](https://www.quantglide.com/). Here's how I modeled performance projections from backtest to live, and and used them to help assess the validity of realized out of sample returns:
* Map out realized returns with the blue lines with data pulled from tastytrade
* Project backtest returns using the historical mean and standard deviation of returns from my 3-year QuantConnect backtest
* Apply an optional haircut on backtest CAGR in line with standard practices
* With the "likely" 15% haircut, the CAGR is 122% at 40% capital at risk
* The shaded green bands show a probability distribution based on backtest vol
* A sample of lines is mapped out to demonstrate the expected randomness in future returns
* The backtest capital at risk is precisely matched to the actual capital at risk of the account for fair comparison
* The realized returns were pretty in line with what the backtest predicted, as you could see
Thoughts?