Quant

Gamma scalping is a core strategy for market makers and volatility traders. The logic is straightforward: buy a straddle to establish positive Gamma exposure, then continuously delta-hedge to convert the underlying’s price oscillations into locked-in profits. You are not betting on direction. You are betting that realized volatility will exceed the implied volatility you paid for. This is the exact opposite of selling volatility. Short-vol traders want the underlying to stay still and collect Theta decay. Gamma scalpers want the underlying to move aggressively, generating enough hedge profits to more than cover the daily time decay. ...

Dispersion trading is one of the few options strategies with a persistent structural edge. The core trade is straightforward: sell index options, buy component stock options. You are not betting on direction or volatility level. You are betting that index implied correlation overstates the actual co-movement of constituent stocks. This edge exists because institutional investors have a structural, ongoing need for portfolio downside protection. They buy index puts, which inflates index implied volatility and, by extension, implied correlation. Dispersion traders earn a premium by bearing the tail risk of correlation spiking to 1.0 during a crisis. ...

Formulaic alpha factor mining hits a wall once the library grows past a few dozen factors: new candidates have decent IC, but correlation against the existing library always exceeds the threshold, so nothing gets admitted. AlphaGPT, AlphaForge, AlphaAgent, and QuantFactor REINFORCE all attack the “generate more candidates” side, while the “the bigger the library, the harder to add new stock” problem still gets little treatment. FactorMiner (A Self-Evolving Agent with Skills and Experience Memory for Financial Alpha Discovery, 2026/02, arXiv 2602.14670) wraps the mining loop into an agent skill and maintains an experience memory that records “which directions kept hitting walls, which templates keep paying off,” so every round the agent samples with a prior built from past wins and losses. On CSI500, the top-40 library reaches IC 8.25% and ICIR 0.77, about 40% relative to AlphaAgent’s 5.90%/0.46. ...

AlphaAgent tackled “LLM-generated factors look too similar, accelerating alpha decay” with factor-level regularization. QuantaAlpha (arXiv 2602.07085) operates one level up: instead of constraining a single generation, treat the full “hypothesis → factor → backtest” pipeline as a trajectory, and apply mutation and crossover between trajectories. On CSI 300 with GPT-5.2 it lands IC 0.0472, ARR 4.68%, MDD 11.8%. Zero-shot transfer to CSI 500 and S&P 500 keeps 40.28% and 19.1% cumulative excess return respectively. ...

0DTE options (Zero Days to Expiration) are option contracts that expire on the same day they trade. In 2022, CBOE filled out the SPX expiration calendar so that contracts expire on every weekday (Tuesdays added in April, Thursdays in May), moving 0DTE from a niche product into mainstream trading. Cboe public materials report that 0DTE contracts account for more than half of daily SPX option volume, with the share continuing to rise through 2024-2025. Public discussion of 0DTE tends to cluster at two extremes: social-media narratives of rapid wealth, and warning labels of indiscriminate risk. This article takes neither stance. The aim is to examine 0DTE as a financial instrument: its pricing mechanics, the evolution of Greeks during the final trading day, the payoff structures faced by the main participant types, and the empirical findings of existing academic work on the realised return distribution of retail buyers. With these in hand, a reader can form an independent view of whether the instrument fits their objectives. ...

Qlib is Microsoft’s open-source AI quant platform, released in 2020 and now sitting at 16k+ stars on GitHub. It bundles data, factors, models, backtesting, and reporting into one pipeline, ships with the Alpha158 and Alpha360 factor sets, and includes 20+ models out of the box (LightGBM, Transformer, TCN, and more). This Qlib tutorial walks from installation to a working LightGBM + Alpha158 strategy, then to reading the backtest report, and ends with a checklist of pitfalls. By the end you will: have Qlib running locally, have downloaded A-share daily data, train a model in 30 lines, run the standardised qrun workflow, and understand IC, annualised return, and max drawdown numbers in the output. ...

This vectorbt tutorial starts from the real difference between vectorized and event-driven backtesting, then walks through the three Portfolio constructors (from_holding, from_signals, from_orders), runs a 10,000-combination grid search, and ends with the freq, signal-shift, and memory traps that quietly produce wrong numbers. ...

The previous AlphaGPT review left an open question: when everyone uses LLMs to mine factors, how long can those factors stay effective? AlphaAgent (paper, KDD 2025) tackles this head-on. Its core observation: LLM-generated factors lean too heavily on existing knowledge, producing homogeneous signals that crowd the same trades and accelerate alpha decay. The fix is three regularization constraints injected into the factor generation process, forcing the model to explore structurally novel, logically coherent, and complexity-controlled factors. ...

Quant trading articles are full of math terms: log returns, standard deviation, covariance matrices, OLS regression, partial derivatives. You can look up each definition, but without the trading context, the definitions don’t stick. This article goes through the most common math concepts in quant trading, each anchored to a real trading scenario with Python code. ...

A stock returned 20% last year. How much of that came from the overall market rising, how much from it being a small-cap, and how much from it being cheap? The Fama-French factor model is the tool that answers this question. It decomposes stock returns into a handful of explainable “factors” and forms a foundational framework for quantitative stock selection. This article starts from CAPM, builds up to the three-factor and five-factor models, and runs a factor regression in Python. ...