Can this be used for machine learning in finance?

Absolutely. It includes specific patterns for volatility-normalized returns and a differentiable Sharpe loss function designed for training Deep Momentum Networks.

What strategies can I implement with this skill?

You can implement a variety of trend-following strategies including Time-Series Momentum (TSMOM), MACD indicators, and multi-scale momentum models that combine different lookback periods.

Does it handle risk management and position sizing?

Yes, the skill focuses heavily on volatility targeting, using ex-ante volatility (typically 60-day EWMA) to ensure assets contribute equal risk to a portfolio regardless of their individual price swings.

How does it handle futures contract data?

It provides logic for creating continuous price series from individual futures contracts, specifically recommending ratio-adjusted methods to maintain consistent percentage returns for backtesting.

Financial Time-Series Momentum

Name: Financial Time-Series Momentum
Author: Donaldshen27

byDonaldshen27

Data Science & ML

Implements professional-grade trend-following strategies, volatility targeting, and multi-scale momentum indicators for financial data analysis.

About

This skill provides a comprehensive framework for processing financial time-series data based on the X-Trend and Deep Momentum Networks models. It equips Claude with the specialized knowledge to calculate stationary returns, implement ex-ante volatility targeting using 60-day EWMA, and generate complex momentum signals like MACD and multi-scale TSMOM. Whether you are building quantitative trading systems or training deep learning models, this skill covers the entire pipeline from continuous futures contract handling to differentiable Sharpe ratio optimization and transaction cost modeling.

Key Features

Advanced volatility targeting for risk-normalized position sizing
Portfolio construction frameworks with transaction cost modeling
Continuous futures contract chaining with ratio-adjustment logic
Multi-scale momentum factor calculation including MACD and TSMOM
Differentiable Sharpe ratio loss functions for neural network training
0 GitHub stars

Use Cases

Creating feature engineering pipelines for Deep Momentum Networks
Developing and backtesting trend-following quantitative trading strategies
Constructing equal-risk portfolios across multiple asset classes

About

Key Features

Advanced volatility targeting for risk-normalized position sizing
Portfolio construction frameworks with transaction cost modeling
Continuous futures contract chaining with ratio-adjustment logic
Multi-scale momentum factor calculation including MACD and TSMOM
Differentiable Sharpe ratio loss functions for neural network training
0 GitHub stars

Use Cases

Creating feature engineering pipelines for Deep Momentum Networks
Developing and backtesting trend-following quantitative trading strategies
Constructing equal-risk portfolios across multiple asset classes