Weekly research digest
| AI Quant portfolio | AI Quant | S&P 500 |
|---|---|---|
| Week to 18 Sep | +0.1% | -0.3% |
| Since inception | +12.1% | +13.9% |
| Since inception, at market β | +44.7% | +13.9% |
| Sharpe | 3.03 | 1.84 |
| Beta | 0.27 | 1.00 |
At market β restates the record at the market's beta — the same strategies held at 3.70× the exposure: the return is multiplied by that factor, the drawdown re-measured on the scaled line, and Sharpe is unchanged by it. Arithmetic on the record, not a second track record — no leverage was used and no financing cost is modelled.
Our strategy development and testing pipeline has produced a large library of alphas: trading ideas implemented in code and tested on historical data. The next question is how to turn that library into a portfolio and al…
The strategy holds SPY long with exposure scaled from a one-day-ahead binomial Markov-switching multifractal volatility forecast. It targets 12% annualized portfolio volatility and exits when forecast SPY volatility exceeds 35%.
we backtested it — Sharpe 0.03 · quant relevance 0.98
The paper presents a daily equity-index directional forecasting and trading system based on OHLCV-derived technical indicators, wavelet denoising and a dual-branch deep-learning model. The reported results cover the paper’s own index universe, and causal feature construction is required to avoid look-ahead bias.
we backtested it — Sharpe 0.89 · quant relevance 0.96
The paper develops a daily volatility-regime early-warning classifier and jointly optimizes probability calibration, feature selection, the decision threshold and a defensive SPY allocation policy. Its holdout results show that complex metaheuristics do not generalize better than random search or a simple ridge-logistic/VIX baseline.
quant relevance 0.98
CAST combines a cross-asset online Kalman predictor with an uncertainty-penalized model-predictive trading controller for systematic equity trading and drawdown control. Its daily price-panel inputs and re-optimization can be implemented with US equity daily bars, although this substitutes US stocks for the paper’s four-market evaluation universe.
quant relevance 0.98
The paper uses conditional diffusion models to generate daily implied-volatility-surface scenarios, applies static-arbitrage penalties and optimizes option hedges from those scenarios. It reports near-zero tracking errors and lower tail risk than classical delta, delta-vega and GAN-based hedges, including during COVID-19.
we backtested it — Sharpe 0.30 · quant relevance 0.93
The paper proposes a method for jointly regularizing covariance and expected returns in Markowitz portfolio optimization. It uses per-mode reliability weights to shrink return estimates and covariance eigenvalue deviations toward a reference state, with random-matrix-theory calibration suggested to identify unreliable spectral modes.
45 new papers written up in this period. Open the Quant Paper Radar →
Backtest and model results are research artifacts, not live trading results and not a guarantee of future performance. Informational and educational purposes only. Not individualised investment advice.