cryptoroddy

Researched in momentum and long-short skew-based straddles option trading strategies; Built strategy pipeline by fetching and processing top20 market liquidity, constructing ITM grid and implied volatility surface, calculating daily straddle signals and 21days rolling momentum signals, applying features cross-sectional regression and combined advanced risk management such as VIX overlay hedge and delta-hedge; backtest proved an 82% winrate and 1.89 SR, with a high Beta1 for skew factors indifferent to market volatility shocks.

Compared traditional volatility forecasting models (GARCH, EGARCH) with deep learning methods (Transformer, LSTM) for high-frequency Bitcoin (BTC-USD) data, demonstrating that EGARCH(2,3,2) performs comparably to deep learning models on intraday forecasts, while highlighting the potential of Transformer architectures to capture millisecond-level order book data.

Enhanced the Shiller-CAPE ratio in stock forecasts by constructing a VAR model with earning yields and inputting stocks by the sum of the capital index, successfully stabilizing CAPE's predictability regarding the adjustment for macroeconomic variation, lowering the RMSE forecasts error by 40% for 10-year nominal annualized returns.